Видео

I typically do not use PLA. However, it's a material to use to test cooling ability.

People who use engineering materials in fdm usually do not post it online maybe that's why you think everybody is printing pla

@@Mekanikalistik You might be right on the perception. Although PLA isnt cheap or better and i genuinlly dont understand the like for PLA. ABS flows better needs less cooling and PETG prints on anything just as clean but flows less. PLA has no use nor benifit. PETG even has less shrinkage and prints more dimentionally accurate.

@@Mekanikalistik It's a cooling test and PLA is bad for cooling. I don't think anything about PLA other than it's great for molds and making die's. I rarely use it, but it's great to test cooling. Do you have some content to share or what is it you would like to see?

@@MrRocksalt Hi mate, you are right! I was trying to explain why pla is very common among 3d print channels. I reckon there is no other better material choice to test your cooling system. But do you really need that kinda strong cooling for such a printer. On that point I do agree with @Nwalker But I also understand the need for your cooling system to handle the fastest melting speed for PLA. But I personally would not waste my time on such a trash material. I can print ABS for fast prototyping too and at end of the day it's all about cost. ABS cost less than PLA in my local market so I would prefer ABS for fast prototyping too. If I need super accurate plastic molds or templates I can always slow down or compensate for abs shrinking, no big deal.

Updated Thingiverse with pic, ~2.2mm. It can go higher a few mm, but try and keep it close to ~2mm

It's all about the correct angle.

@@MrRocksalt yes I understand later how have you done, but it doesn’t make any problem with the filament path angled ?

@@occasionalriders1885 The curved path happens after the filament guide. The angle is ~2.6°. Probably does not matter.

@@MrRocksalt mh understand... Have you ever think about the nf cannon extruder Gear?

@@occasionalriders1885 These were parts I had, except for the K1 gear. It's size is almost perfect for this extruder, for the configuration I wanted to use it in. But I need to wait for the SLM parts so I can determine it's function.

Polyetherimide

LDO 1004AH.

@@MrRocksalt 14HS10-0404S Could give you 40% more torque in about the same foot print if I’m not mistaken. 9Ncm vs 14Nmc.

@@1943vermork 180grams., no thank you!! This is the one: CSE14HRA2L4100

Endcaps on nema 14 are very low weight. ~3-4 grams guessing. To me, better would be a custom motor, lower profile, but wider.

@@MrRocksalt high torque pancake motor 🤔

Print it and give it a try. It is not as good as the valve duct, but it is sure good IMO. There is a blocker in the front so the 2 air front's don't collide and a splitter in the rear. The rear ports are offset from the air source and a small divider that helps those 2 ports not get direct air. It also has 2 valve slits in it to re-direct air to the sides. www.thingiverse.com/thing:6779529

Speed benchy is a way for me to see cooling challenges. Nothing more. Just a tool.

I really would like some specific tests to print to compare worm to others. Specifically, models that show some range of artifacts that can easily compare 2 extruders. Often, I will see artifacts related to a specific model that may or many not be seen to others. Do you have any you use in comparing extruders?

I use them both and have not made the newest version UFO duct for the chubby yet. Too excited to try this extruder and used what I had.

This is Goliath and can flow 92 with 0.4 CHT, I don't care for the print quality with CHT so I use tungsten carbide nozzles and limit my flow to ~65-70 in slicer.

It's smooth PEI. It looks like that for 2 reasons. I sand the PEI with sandpaper to scuff it, to make more valleys for the filament to sink into(like painting on a smooth surface vs one with scratches). Sticks like mad after cleaned with dish soap and a green scrubby. But after some time I am lazy when it starts to not stick again, and put some glue stick for some material that never sticks to anything. That residual glue stick makes a whitish color to the bottom of the print and also on the plate.

I’m so interested too in what PEI is that. It looks a modified sheet with an adhesive Buildtak PEI or something like that. Can you please tell me what kind or brand of PEI do you use please? 🤗 Thank you so much my friend! 🤗

@@guzmangomezguerreiro3418 Gizmodork PEI. A lot has to do with how you prep it.

@@MrRocksalt Yes I know it, but im curious because your plate looks good for me. I’ve a VzBot and I don’t get any decent PEI or spring steel plate. Do you scratch the spring steel first and then apply the PEI or something like that? 🧐 Thank you so much for your reply!

@@guzmangomezguerreiro3418 I believe PEI is PEI, unless whomever sells is selling something else! If you have issues with the glue, welcome to the club! I have much better life from the glue with temps <106C. Yes, I sand the spring steel, bed plate, and the printing surface.

I think it's just too fast. I pumped up the acceleration on this one and the square corner velocity. I think the previous ones were 7 minutes vs this at 6. Probably overheating, but the top of the model is not as cooling sensitive and still there. I need to make a few changes back to where I had the extruder acceleration set. When I watched the part print, it seemed like there was a stall at each layer change. Maybe that stall was enough to overheat the plastic.

it's out in beta

www.thingiverse.com/thing:6779529

What is that? The stock duct? Have not tried Valvnator on VZbot?

@MrRocksalt I'm talking the stock vz duct not valvnator

@@WickityWack Yes, I understand, but I don't have a VZBOT or use his duct. No way I would know. But from reports, the valve duct is better buy a good margin and this UFO duct is on par with the valve duct. Still beta testing, but so far it's rocking well.

@@MrRocksalt You dont?! I really thought you did. But anyways if that's the case ya it probably is better

In many ways, yes. For raw cooling power, no. VaLvNaToR flows ~115gpm and this one ~85gpm. They are very close, but this duct has better 360° coverage and more pointed/direct to nozzle cooling. It does not breathe like the VaLvNaTor and has more losses. Prints 8 min PLA benchy, have not tried faster, but it's clean, cooling wise.

Thanks for sharing!

My logic was to show the difference in pressure seen at the part relative to 3 of the same ducts. All flow of the ducts are similar GPM, same nozzle sizes. Having high pressure, at the nozzle orifices does not mean it's better. More velocity might cool faster, but it may cool too well and layer adhesion issues. You talked about Berd Air, and they are OK, not good, not great. I would need at least 3 berd pumps to come close to what this does, maybe more.

@@MrRocksalt after thinking more about it, what this scale approach does is measure the power the airflow is capable of delivering downward. As a comparative with similar nozzle locations, exit angles and sizes I can see how it would be good for a proxy of overall cooling potential. But for comparisons between varying nozzle sizes, locations and in particular directions, it may be misleading, as this type of power can be “wasted” doing things that could help cooling like turbulent mixing, flowing laterally, etc. I’ve been thinking of a way to measure this better and I think I’m into something . The idea would be to create fixtures that hold a thermally isolated small SMT resistor(an 0201 sized one is close to the line width of a 0.4mm nozzle) and a small glass bead thermistor tightly thermally coupled to it from underneath(cyanoacrylate works well at these size scales). Using a lab supply/or well sized resistor/power supply and a DMM to measure voltage(power), find a power level that doesn’t melt the supporting material with no air flow, and put the fixture on the bed. Slowly grid scan the area around the fixture moving the print heat, and note the temperature the thermistor reads. Knowing the delta between the ambient fan output temp, and comparing to the delta temp with no forced air(and maybe covered with a bit of insulation to eliminate natural convection, this should produce a mapping of effective thermal resistance C/W. Sort of like a bed map, but for cooling performance. The trick will be to make a thermally isolating fixture that won’t melt-some sort of foam maybe? This size of resistor, should have a bare thermal resistance near 200C/W so it shouldn’t take much to power it. Changing the shape of the fixture around the resistor could be used to simulate various printing conditions. A large flat disk with the resistor in the middle simulates direct on-the-bed printing. Putting a resistor on the edge of a large plateau simulates walls on a large print. Other scenarios could be interesting like on a thin wall, on an overhang, etc. Knowing how much cooling needed is going to be a balance-fan output temp, difference between filament print temp and Tg of the material, the plastic’s heat diffusivity, and how fast the head is moving around could be simulated in LTspice or similar, to understand how quickly the filament is likely to cool.

@@brandonhicks7549 This is an early stage of this duct and I think I have about 3 weeks into this shaped duct. I have changed it so many times for some reason or another, based on the original I started. I don't know your methodology in creating a duct, but the methods you mention to test come later in design, at least for my workflow. I am not a fluid engineer by far and, unfortunately, I do a lot of trial and error. If not for CFD it would have taken me much longer to get it this far. Possibly another test bed is to use FLIR camera to see the effects while printing. It would not be as comprehensive as you mentioned, but an overall view of what is happening.

ABS, these are super tiny valves and can overheat easily when printing. IIRC, that was 60% fan.

This is a new concept duct for me and Chubby takes up the entire center of the UFO ring. In fact, I had to model the UFO duct to fit the Chubby. Then I was thinking about the ability for the duct to en-train air form the top and wondered if there is impact with Chubby vs Goliath to that air. When I pass this duct along, I want to make sure it works with Chubby or Goliath or or or. Hence the change to Goliath.

Oh okay, I understand it. I’m thinking about replace my goliath on my VzBot and try chube hotend but I don’t know if it’s worth it or no. What do you think? Nice job that duct and awesome printer. So impresive with it, I would like to do something similar 👏🏼

@@guzmangomezguerreiro3418 they both work well. Is there a difference in printing, I don't think so.

I sent the original UFO duct to him to test. I hadn't been working on it very long with that version and it does not cool a PLA benchy past ~9 minutes. The valve duct(VaLvNaToR) can print PLA benchy in 7.5, but the package size does not fit for the template he was asking for. Since then I have improved the duct and the entry timeline is over. But.. I have learned some things along the way with UFO duct that I had not seen before. Layer adhesion becomes a problem with the air so concentrated at the tip. How to test fairly a duct? A speed test is not it IMO and also dependent on material. It's a Pandora box.

Or maybe high speed cooling is not the right idea with regard to layer adhesion? I can cool well with 5 grams force vs 20-30. Which would you prefer?

@@MrRocksalt It looked to me like you're still chasing faster times than you can currently get with PLA. I don't doubt you can get very good results in normal printing but I do think the uninsulated nozzle is a big performance detriment and fixing it could unlock a lot.

@@daliasprints9798 I do understand what you are saying. What I don't understand is your value of 40C below set temp for a nozzle that is not insulated only on the sides. I do believe there is heat being taken away, but what is the actual value and impact? Have you ran tests with a thermocouple at nozzle level to verify when printing? Probably a hard test to conduct considering not all ducts are aimed directly at the nozzle and it's velocity would have the greatest impact. It's something I need to look into as the sock has been trouble for most hotends.

@@MrRocksalt Granted IR imagery is imperfect for metals, but I get significantly lower readings on exposed nozzle surface than block temp with fan on at any appreciable level, but do read near full block temp without fan. I think with some math you could infer how much cooler the core of the nozzle is going to be with all 7 faces exposed than with just the bottom face exposed.

It's no good right now. Too much velocity. Taming it down, testing. Will share this one, but I want it right.

My intention was to make a generic bottom end to be adapted by the user. If I can't get this duct to work like I want, no point in sharing. This one is no good right now.

​@MrRocksalt what's wrong with it?

This cube is not meant for vase/spiral mode. Try one and see what you get!

Chubby (Chube)

Supports?? 😂

@@daliasprints9798 no, none.

What is the reason?

@@MrRocksalt It is more watchable, and with a little bit of additional initial set up, actually more filmable(if that's a word) than what this is that you did.

@@dandcc9192 It being a 22 minute print, I didn't bother to post the whole video. Instead I posted some speed and a follow up video of the final product, listed in the description. Just playing around.

@@MrRocksalt "It being a 22 minute print, I didn't bother to post the whole video" you don't notice how it sounds like a problem a time-lapse would have solved?

@@dandcc9192 No, not really. With time-lapse you are not getting the feel/look of the speed vs quality. I guess some prefer your idea and others don't. Don't assume everyone has the same viewpoint.

Some others will be testing it soon. So far, I have had to decrease cooling % using this duct, and by quite a bit. I don't have a lot of time on it, but from some speed tests it seems to have better overall coverage/°360 air. Time will tell.

Best one I made to date IMO!

@@MrRocksalt Nice! I saw your benchie video for another version and you sounded over the moon :)

Yes, They sure are powerful motors moving this much mass! But they do resonate in the 75-175mms range.