The problem with the constricted air duct was that there was too much constriction (it needed to be wider by about 50%... roughly half way between where it was and the edge of the hole)... also, it should have been printed so that it stuck into the tube, not so that it was 3 or so inches from the opening. Being inside the tube would have drawn in more air instead of just ramming into the cross-currents present in the gap.
Lots of static pressure. the fan was built for volume, not static pressure. And the noose was too short and didn't drop air inside the tunnel to draw it back in. But the neck is too small and thus turbulent. Vanes might have helped.
@@kristmadsen What Is Islam? Islam is not just another religion. It is the same message preached by Moses, Jesus and Abraham. Islam literally means ‘submission to God’ and it teaches us to have a direct relationship with God. It reminds us that since God created us, no one should be worshipped except God alone. It also teaches that God is nothing like a human being or like anything that we can imagine. The concept of God is summarized in the Quran as: { “Say, He is God, the One. God, the Absolute. He does not give birth, nor was He born, and there is nothing like Him.”} (Quran 112:1-4) Becoming a Muslim is not turning your back to Jesus. Rather it’s going back to the original teachings of Jesus and obeying him
The Dragon’s Breathe fan was designed for flow when it needed to be designed for static pressure and allow the flow to come from Bernoulli’s effect. It might perform a lot better if the restriction had a wider outlet hole to reduce the amount of pressure the fan is trying to fight against.
Or, you know, not have a restriction at all, and attach the fan directly. Makes no sense to attach any restriction, any restriction will make the fan more less air, and for this test you want it to move as much air as possible.
@@jort93z That's ... not how bernoulli's effect work, Think of it as tunnelling air by creating difference in pressure, as the comments said if the tunnel was wider, it would have performed significantly better.
@@amineabdz Tunneling the air is not the issue, you can get that to work, the issue is moving more air than you started with. They want to get more airflow with this device, that part won't work. Any flow restriction, doesn't matter if it is 10% or 90%, will decrease the amount of airflow. If you want this to somehow get more airflow, you need the area over which the air is spread to be considerably larger than the fan. The measurement tunnel would need to be wider, and in that case the measurement isn't comparable anymore. just attaching it directly will be more effective, no matter how they redesign their "Air maximizer".
@@jort93z you still don't really understand this, volume stream is velocity*area, if he made that restrictive tubes diameter about 10-15mm smaller than the air tunnel and made it go into that tunnel, he would benefit from that restriction. Air speed would increase by couple % and the pressure difference would suck more air through that 5-7.5mm gap between restrictor and tunnel. Dude said he studies mechanical engineering, I guess he should go back to drawing board and pay more attention at college
@@mikoajkosma4010 Well, but you are forgetting that a restriction will make the fan work against a higher pressure, making it work less effectively. If you could just make a restriction and increase airflow, surely you'd just decrease it by 99% and get fat gains, but that's not how it works. The fan will need to work harder and move less air. Well, when experimenting, some experiments go well, others fail.
The problem with making a constricted air duct is that the back pressure from trying to compress the air fights the already low-torque motor. I had this issue when designing Ov3rdrive as well. But the flow on the batwing was beautiful!
yeah. it is restricting WAY too much. it should have maximum 20% restriction or so if even and the gap to the duct is way too big too. the end of the restriction should be inside the duct or ending directly at the duct with around a cm ring of gap for the other air around to pull in. (the values are picked by my gut and are most likely wrong or inefficient, but should work better than what was in the video)
Correct, my main complaint about Noctua is that my standard dust mesh on my computer is enough to completly hamper the airflow on the noctua fans, even though they are SP fans. I was very very surprised especially because i bought the 3000rpm pro fans from Noctua, they only states they have a sp of 7.6mmH2O where my current fans has 65mmH20 difference was 15C in case temperature. On top T30 also only have a sp of 3.3 which is completly absurdly low...
Yeah, I think the funnel shouldn't decrease in diameter as aggressively as it did. I wonder if a secondary venturi tube for the outside air would help...
No flow restriction will help it flow more air, that makes no sense. It will increase the velocity, yes, but it will decrease again once it enters the test-pipe.
@@jort93z the idea is that it creates a low pressure area that is taking air from the gap with it into the duct like a "fanless fan". the problem is just the way it is shaped and that the idea doesn't really work on such low pressures and low airflow.
A low restrictive venturi tube would be more effective. Since it's a fan and not compressed air aiming for a 5-10% increase in airflow may be achievable.
@@Selbitschka but it was a very clever idea. Tweak the design and submit again, I think I've seen some comebacks in the series. These low pressure "fluid grabbers" when tweaked right are very efficient. I use venturi tubes to inject ozone in my house water. They suck hard (pun intended).
@@Selbitschka Try measuring the pressure produced by the fan. That will tell you how much you can narrow it. Use a length of vinyl tubing with water in it to measure pressure.
Hey man thanks so much for showing off my fan (Blowsie, Blowie, whatever)! I can't believe it did almost as well as the stock design. I designed it to suck air into the center so I was very happy when the 2nd smoke shot showed exactly that. The low RPM is very interesting, I honestly have no idea how thick the blades are so I can probably lose a lot of weight and get some better performance if I load the design into some beefier 3d modeling software. Anyway thanks again!
I've been 3d printing gaskets :) Works good once you get the hang of it. So far I've put them on >> rainbird anti siphon valves ; Water hoses ; Lots of water pumps ; Sinks/faucets . I got tired of buying them. They break so often it seems like a conspiracy ! For high pressure stuff, I will include a space to fill with aquarium silicone or E600. PETG
I think it has more to do with the static pressure and that's why it goes out the side, if the pressure was lower the fan would pass the air right by the air gap and out the back like it should. But since the end of the tube is so small the air isn't compressing because the fan is weak, so it just increases pressure and flows out the sides.
in the rc airplane world when we use edf motors, there's something called FSA fan swept area(physical area swept by the blades minus the center hub area). how you neck down the tube behind the edf affects your top speed\ max thrust, a thrust tube that is 100% of the fsa will have higher static thrust(torque) and one that is necked down to 90%fsa will have higher top speed(horse power) after a certain point when constraining the thrust tube it becomes inefficient, and unable to push past the squeeze. the max we typically use is 80%. that Bernoulli tube looks like its 50% at best. id put money that the multiplier would work better if Different Fsa diameters were tried. if the fan is 120mm and the hub is about 40mm then that means 100% fsa would be an 80mm exhaust tube, and 80% of that 64mm so the workable range from 110% fsa down to 80% would be about 90mm to 65mm. that JUST the tube being efficient at its job at speeding up the air. this doesnt take into account the distance and area available for the that sped up air to "drag" more air along with it(distance from end of the tube to the beginning of the next "orifice" id start with an 80mm tube and see what it does, then id go right down to 65mm and see if its any better. it may still not work, the motor may not have enough ooompf to speed the air up significantly. only then maybe start adjusting the distance from the end of the thrust tube.
Your FSA calculation is wrong, you can't do 120-40=80 , you need to calculate the areas before subtracting, the 100%FSA exhaust tube is 113.1mm, 80%fsa is 90.5mm
Huh this is something i did not understand correctly even though ive built thrust tubes in the past. I have some edfs lying around should build an aircraft
You should really look up the video of the cheater on here. It has smoothed out air flow behind it because it has like rear cone pod behind the motor. The cheetah fan on that cheater frame would have the best chance of beating it.
The ideal angle for a duct like that is around 11° of divergence/convergence. Just needed to be more gradual. Also having the exit closer to the wind tunnel exit would be better.
The dragons breath needs a second slightly tunnel to funnel that air and any extra air the fast moving air pulls in. The problem is that with all the extra open space now suddenly the air runs into turbulence and rapidly slows down and the pressure increases blocking airflow
It's kind of ironic that the idea was to reduce the area but then overlooked the sudden increase from a finite area to an "infinite" area at the end (which to be fair for liquids this wouldn't matter, but this is air which is compressible which always threw me off when I took fluid mechanics). I'd be more curious on a test where this design was mirrored so that it would be a Venturi tube instead of half of one, just to see if it does anything.
I bought Anemometer and made a test rig, I got Bernoulli's principal to work but I used a blower style fan design. The best I got was 660ft per min after 7 rotor designs.
Hey @TheRattleSnake3145 did you try any of the principles @RichardInTN (top comment) recommend like adding a shroud so the air is being sucked in, or changing the fan to be a compressor/static pressure style? Also I like the idea of creating the vanes on the rear duct, no Idea if creating a vortex would increase flow though.
@@mavric1177 I didn't use a shroud for the blower design but when I gave up on the blower/Bernoulli design I designed a more traditional design. The shroud infront of the fan made a significant improvement to flow. My current best design beats the noctua and I believe just beats the cheater. However my actual individual results will be different to his as my test rig is different. My first try at veins behind the fan reduced the airflow.
Venturi ejectors are like a class 3 lever for airflow. They convert high pressure, high speed, low mass flow into slow, high mass flow (pressure is torque and speed is MAF in this analogy). Axial fans are already slow, low pressure, high mass movers so it's like trying to start a bike in high gear.
That's the assumption a had myself also if it has streaks inside it would make more acceleration i believe , but overall great Idea , the implementation need some more work 💗
Just too steep a funnel for the end diameter, the air is being bounced back out before it can squeeze down that narrow throat. The spacer is also too large, the end of the funnel needs to already be inside the wind tunnel or boxed in. If I was to refine the design I'd drop the spacer so it's only as tall as the funnel and wall it in so it only has 2-3 cm of gap where it meets the fan. The throat of the funnel would be ~20% larger with a straight taper and no neck, maybe do some testing to see if vortex generation has a beneficial effect but I'd probably keep it smooth for simplicity's sake.
@@Tgspartnership it will give it an extra chance. In the smoke test I saw a lot of air escape through the sides where the funnel connects with the fan.
I'd like to see the anemometer placed right at the end of the cone shaped thingy. I think a big part of the problem is that not only is it constricting the flow of air going through it, but as soon as the air gets to the end of the cone, It just dissipates out in every direction because it's no longer constricted. But the air coming right out of it might have some speed
I think if the constricting tube was halfway placed inside the air tunnel it would help out a bit. (Plus a bit o' size increase of the end of the funnel could likely help)
You should try more testing with the t30. It would be cool to see fans that would be more efficient with more rpm/ torque. Fans like the contra rotating one and the jet engine compressor style fan from about a year ago. Also I’ve seen this a lot in your yt comments. But try making different fan classes. One would be a standard class that it basically all the fans that fit in the frame of the a12x25 and ones that have external components outside the frame of the a12x25. The external class would be interesting with fans that don’t work out like the dragons breath.
I've noticed with all of the air flow modifiers like this one, the Dyson style, etc., that they would work if the motor fan produced a significant amount of power like a vacuum motor. The challenge still remains to create a low power fan to increase it's airflow as what people have been trying. I think if this dragon's breath adapter has the same outlet diameter as input, then introduce the venturi effect they are going for, it may work! Or make the outlet diameter size as if the fan hub was not present. So just blade diameter with no hub. Just a couple of ideas. Love the series!
i think the cone would have a better chance if the smaller end is inside the wind tunnel, with only an inch or so of air gap to suck in air. maybe even better to add fins on the outside of the cone to align outside air coming in.
Bernoulli requires pressure potential to be exchanged for kinetic potential. Fans accelerate air flow, they do not pressurize air. This gets into a deep engineering field of fluid movers; pumps, compressors, and fans. In compressible fluids movers are separated into fans and compressors because they add energy to the air in fundamentally different ways. Fans do not compress air, they simply add energy to the air in an incompressible fashion, similar to incompressible fluid pumps. Compressors on the other hand are moving with such speeds that the fluid can accumulate pressure potential locally without necessarily accelerating (it happens both ways) meaning the fluid can come out of a compressor with the same velocity it entered but at higher pressure. With a fan you are never going to get that compressible behavior because they simple do not move fast enough. The whole point of fans is to add small amounts of energy to a lot of air. So back to this design, if fans simply increase the velocity of the air, then any area reduction will further reduce the pressure, and now we run into a conflict. Fluid will flow from high to low pressure, and if all we did was increase the velocity then we are now below atmospheric pressure. Meaning are total flow must reduce to keep fluid flowing in the correct direction. The whole goal of this exercise is to design fans that maximize the energy exchange efficiency between the motor and the air. This means low turbulence, one dimensional airflows along the axis of the fan. Just some tips if you are going to investigate along this path, constant flow areas and flow stabilizers are going to give you the optimal performance
I'm not exactly an expert but I thought he had the airflow backwards! Pretty sure the idea of speeding up the air coming in would be an excellent way of making a better fan.
Woohoo, new Major Hardware vid! Looking forward to seeing the new fan mount in action. Always an awesome evening view when a new Fan Showdown episode drops. Thanks for the awesome content and constant improvements!
I think the best part about this video is how he doesn't even realize the pun when he repeatedly said he's a "BIG FAN" of the new purple print material.
With the "Dragons Breath", maybe a suggestion is to shorten the "legs" on the funnel shroud, so the funnel is ending inside the wind tunnel? I am not sure of the physics, but that feels like it might be more effective.
If you look at the outflow of all the fans, most of them produce a significant amount of radial outflow. If the dragon's breath also had this issue, then the radial outflow from the fan would hit the channel wall almost perpendicular, leading to more collisions with the axial outflow and thus a significant decrease in exiting flow velocity and a resulting increase in pressure, meaning it had to fight even more static pressure within the funnel. A channel converting this radial outflow to be more axial could have a better effect than just a normal isentropic nozzle
The size change of the funnel was probably too much too quickly, so the air would rather blow out the sides at the front than go through the funnel. A less severe funnel might work.
I'm super glad I started watching this series since either near the end of S1 or the beginning of S2; always satisfying to watch the smoke flow through the fans and the rad 80s-style music you picked starting from IIRC early S3 (thank you again for identifying it as a Mattie Maguire stock music track) has been a great choice for the showdown montage! :)
Wow! Over 200k subscribers! Congrats Major Hardware, your hard work, consistency, persistence, and viewer engagement is a model for creating a community on UA-cam. Here's to the next 100k!
I feel like Chad did a lot of fluid dynamics math and just completely ignored the fact that CFM through a fan is totally different than CFM through an air pump. This is a really good object lesson in what backpressure is and how it works (and in the case of an axial fan, how it makes things *not* work).
@ Major Hardware add some some tiny holes on the funnel part of the dragons breath to help alleviate some of the drag inside like in the engines of the sr-71 blackbird
Your videos are excellent at displaying the airflow capabilities of different fans. This is why I'd like to request a video demonstrating different CASE airflows. Idk throw in a cardboard blockout to represent the general shape of a gpu and ram, for the air to swirl around. Naturally, I'd like to see the most air possible dragged across the gpu. --Then maybe compare fan positionings for airflow. 3 fans vs 2 fans, 120 vs 140. Using a vent for either intake or exhaust. Or just displaying how different cases, in general, respond to the same setup. --Like, do their more-or-less-the-same dust filters change the airflow in the brands' case? --do the cases that only have vents on the sides of the front panel, surprisingly, actually pull the same amount of air? --does the latter change how the air is flowing internally, because of the angle the air is being dragged in?
Fan designers need to keep something in mind: Virtually every fan design does not merely push air in one direction. It also pushes air outward. When that air moving outward hits a sloped wall, it is "reflected" by that wall. A straight wall will have little effect apart from some friction, because while the air might "bounce back," it's still moving in the direction of flow at roughly the same speed. If the funnel constricts, the air will "reflect" back toward the fan. If it widens, it will "reflect" in the direction of flow. Nobody, from what I've seen, has taken advantage of the latter, and many have fallen prey to the former.
there really should be a catagory for fan add-on prints that just get testing with a stock a12 as well as whatever the creator sends in. i'd love to see how this and the dyson perform with a standard fan. that would actually be useful for someone who doesn't want to buy all new fans but wants better airflow and has a printer and doesn't mind a unique style flair
The main issue with the dragons breath is vortex that develop in the funnel, I ran into this same issue when trying to make adapters for 40mm noctua fans for my 3d printer. You can see the issue in the smoke test where air is pushed back out the fan. You can mitigate the vortex by adding stator veins in the compression chamber, but over all you will lose output not gain. (the part cooling adapters i made did not produce enough static pressure after adding those veins to cool PLA so i gave up on the idea and went back to using dual 5015 blowers) (the info i learned this from was some HVAC documentation I randomly found so yeah apparently this is a known thing in certain circles)
to cool pla with 40mm's you can't(or don't much need anyway) constrict the flow, instead an open air guide works well enough. needs insulating the heatblock tho. that being said I have one like 10watt 40mm delta I can swap if I need more cooling.. that thing blows. downside is that it sounds like a vacuum cleaner.
@@lasskinn474 Yeah i was going for a silent printer :P As for constrict vs open, depends I guess? I was aiming more for a general purpose design as I print other materials as well. Either way the experiment wasn't worth over using 2 good 5015's.
@@nobodytoyou4887 yea i just cut power to the fan when printing abs, but a constricted highly localized constricted one would be more useful with abs I guess if printing small bits, so as to not cool the entire piece but not have it mushy up small peaks.
There seems to be a lot of air coming out from the sides. Maybe using the "radiator gasket" or whatever Noctua calls the gasket that goes all around the fan to seal it to radiators could be a good idea for testing in general? Less losses!
It's quite clear that there can't be any openings between the fan and the air intake, or else most of the air will just escape from the openings. The idea of a longer tunnel could maybe work better if it's completely sealed. Perhaps it could have two fans, one on each end, or something.
I really do love this series! And I'm a biochem engineer who's never built a computer in my life....you just make it very interesting, the good old trial and error engineering really speaks to me
@@pirojfmifhghek566 haha, yeah that won't be difficult to print at all! Also, if the tips hit the casing they'll turn into ADP which will deliver far less energy ;)
Perhaps the 'Dragons Breath' would have worked better with the wind tunnel if the 'flow maximiser' had a larger tube diameter and lesser gradient, and for that tube to protrude into the Wind Tunnel for say 10 cm, with a clearance of say 1 or 2 cm between the inside of the Wind Tunnel and the outside of the 'maximiser' tube. This should allow for the lower pressure to be generated at the wind tunnel opening. I guess I'm saying like an inside out Dyson fan. Sorry I don't CAD, so I can't draw it.
Bernoulli and laminar flow seem to go hand in hand, im not sure if that design managed it there but it was still a brill design and concept, honestly a good fail can really beat a poor win. I signed up, great series, looking forward to seeing more contendors 🙂
Watching today's episode just seemed to reinforce my idea that the smoke needs to be more uniform. What about a nozzle that attaches directly to the smoke machine and creates a curtain of smoke instead of a linear flow, that can then cascade down in front of the fan to provide a more even flow that demonstrates the fans' airflow with complete independence from how you squeeze the bottle, or where it pours out?
Here's an experiment I hope you'll do: The bernoulli fan's funnel-and-supports, and make the supports half the height, so the funnel sticks a bit inside the wind tunnel. Alternatively, extend the funnel with a pipe part which is half the height of the supports. Getting the smaller pipe inside the bigger pipe would increase the chances of the funneled air to pull along the outside air, because the current configuration makes high pressure air which then spreads in all directions instead of continuing straight through the wind tunnel. You could also test those changes with fans with a higher angle of attack, to get more static pressure at the price of less airflow, to deal with the loss in efficiency caused by the fan needing to compress the air so it fits into the funnel.
>Be Chad. >Walks in. >Yells with an open chest: "Cheater beater." >Fails even in being registered by the sensors in testing. >Refuses to elaborate further. >Leaves. What a Chad.
You should try it again without that massive plastic frame structure! Let the bottleneck portion enter into the tunnel rather than let that plastic stand set it an extra foot away -letting all the ambient pressure out.
The bowl of the funnel from that multi-part fan was the problem. The bowl didn't have enough depth to it before it got smaller. That created a buildup of air that then slowed/stopped new air from being able to enter.
S5 should use a Delta fan as the hosting unit for all the blade/propeller designs! They are insanely string units usually used for industrial server applications. It would be cool to see an all-out, as wild as possible, competition!
Would love to see a bonus episode where you take some of these fans and pair them up with other ducts people have designed. Maybe even specifically testing other fans with the cheater duct? E4 Mafia!
The thing about the Bernoulli Multiplier is that the shape was meant to take a high pressure air stream and convert it into a high volume air stream. If you already have a high volume air stream it won't do anything. The shape does work on emergency escape slides.
I'm so happy for a non-working idea. Those can turn out to be the way to learn. And I've already seen some comments why it didn't work and how it could be improved. I hope to see the continuation of this design.
9:10 idea was sound but he reversed the tunel, air is fastest near blade edge, so if you would do that and make an outside ring that is the push (like dyson is doing) then you can pull much more. problem is that this depends on PRESSURE and you seen how easy it is to leak some. I am sure it would be much faster if fan would be duct taped to the tunnel, stopping the precious pressure buildup from leaking everywhere...
I think you almost need to mount the deagon's breath backwards so the fan pulls higher pressure air from the immediate front of the cone, allows it to expand in the fan and go out the back
With standard bladed fan designs, venturi tubes don't work on output, only input, they simply do not have the pressure capabilities to push the air through a restriction and it will just bleed out the holes between if not sealed well enough. But, if you truly want to use a venturi tube, you have to put the fan half way between the intake curve and the output hole and seal the fan housing properly.
Bernoulli's principle also states that with a decrease in static pressure comes a rapid decrease in temperature. If you give that air room to expand before it hits your heat sink then you effectively have the standard cooling from the air flow over the heatsink, but also some adiabatic cooling from the rapid expansion of air You could potentially improve the Dragon's Breath adding ANOTHER tapered mouth on the sucky part of the fan. You would also need to seal the edges otherwise that pressure is just going to escape.
Just for grins... Take a couple of cuts at the maximizer... Start wit an inch, then 2, then 3. I think what the problem we're seeing is too much pressure drop in the tube in its attempt to speed up the air to drag more in with it.. By cutting off progressively more, you'll reduce pressure and get more airflow through it which should gain some numbers.
Bernoulli's theorem just says that you're basically concentrating the pressure into a smaller volume, but it adds drag so you'll only ever lose pressure from it
Thank you so much for featuring my Dragon's Breath fan design. I am in the process of redesigning it now and am actually going to do some testing to see if I can get the desired effect I was going for. I hope that the redesign can be given the chance to redeem itself. :)
One of the difficulties with putting a converging nozzle on a fan producing substantial pressure is that the output flow (from the blades) is (usually) quite rotational. The bits of air move in a helix. Reducing the diameter of a rotational flow means squeezing the rotary motion down to a smaller diameter, in a direction anti-parallel to the centrifugal force of the air. This work, basically the opposite of what happens in a centrifugal blower, produces a force that opposes the forward flow of air. I am not sure how big this effect is, or how hard it is to remove the rotation from the flow with a set of stator blades or similar.
Suggestion for the fan supporting wall: Either use some tape around the fan frame's perimeter to keep air from leaking through the gap between fan and plexiglass, or get a watercooling radiator fan gasket so it seals correctly. If you look carefully, during the side-on views you can see smoke sneaking out from between the fan and the wall.
I dropped out of this since S2 or something, but wow, I'm amazed to see it's gone on to season 4 now. People are really keen on their fan design skills haha _Also the batwing smashing the smoke test 👌🏻👌🏻_
The problem with the bernouli fan is that the frame was designed to be surrounded by the wind tunnel not open air. The concept is that the air in the tube already creates the pressure and the air coming from the fan is accelerated by it.
I would love to see a test that shows resistance to recirculation. An example of what I mean are some mid-cooler fans by Zalman: they initially shipped with plastic rings around the outside, but overclockers discovered they are more effective with the rings removed. My rudimentary testing with little strings stuck to the interior of my tower suggested this is because the air spilling out the side of open fan blades creates a barrier that reduces recirculation of warmer exhaust air back around the fan to mix with the cool inlet air.
i think there is clearly leaking pressure of the dragon breath after seeing the smoke test, prob it wouldnt imrpove much but there is a gap between the noctua and the air duct on atleast 2 sides
The end of the adapter needs a bell exhaust for it to work. Also the narrowest portion of the throat is too close to the fan intake. Small changes should big a very big difference in the end test.
I think a lot of designers seem to underestimate how impactful back-pressure is. If you have too much back-pressure it creates a turbulent flow before even passing the vanes
The cone needs a gradual longer change in size to allow air to funnel and not abort as it gets compressed since it’s an open air design. Or you can have a longer large volume of high pressure low velocity necked down to generate low pressure high velocity (which your design has) by calculating the size of tube needed to optimize your output with the fan size (Venturi effect). Just my $0.02
It's funny, if you put the flow maximizer on in inlet instead of the outlet, it would at least increase pressure but in it's intended use it appears to be a flow minimizer. lol
One problem with the "flow maximizer" is that it makes the tube much longer than needed. Also, the smoke test shows that the air comes out of the back in a 45° cone, which can't work with that gap. If it was only leaving half an inch or less between the fan housing and the wind tunnel edge, it would at least spit out its air partway down the wind tunnel, where everything COULD be redirected forwards.
love the plexi fan setup as it separates sides…that dragons breath is thinking backwards. he was try to make a velocity stack, those work on suction. that fan isn’t the right type to create static pressure or act as a compressor…
I think if you want the Bernoulli tunnel to work you need to run the end through the mount. That way you are pulling air in the tube. This one has too much venting.
I'm watching smoke pour out between the fan and the "duct". You have to at the very least make the connection airtight to give it a fighting chance. Any air pressure being built up just flows right out the non sealed joint.
If you were to put some duct onto the front of the fan it would create an area of low pressure which would help drive air through the nozzle. Referring to the constrictor nozzle thing.
I'm at 5:03, haven't gotten to the tests yet. I'm guessing the 'cheater beater' will fail. You don't get any more air volume through, compression of the air by the reducing cavity increases flow rate at the smaller diameter. But if anything, your going to reduce overall air volume through the system by the resistance of that compression.
Holy FRICK. This particular video was a mind blown moment.... the reason why turbine inspired fans don't work as well is because turbines have to compress the gases. They're not just moving air fast, they're compressing it. So the fans without housing/flow modifiers are going to do better if they have some kind of progressive pitch/lift generating shape with low to medium blade spacing rather than flat pitch high blade count.
You can buy 120mm rubber gaskets for the fans on amazon for cheap, removing the corner rubber is necessary for a tight seal. I believe you will see a increase in static pressure when eliminating the gaps. I did this for 5 radiator fans and dropped my average water temp
Dragon's Breath expectations: Feels like that's just going to make a lot of unnecessary resistance, then turbulent air escape before it enters the wind tunnel.
The problem with the constricted air duct was that there was too much constriction (it needed to be wider by about 50%... roughly half way between where it was and the edge of the hole)... also, it should have been printed so that it stuck into the tube, not so that it was 3 or so inches from the opening. Being inside the tube would have drawn in more air instead of just ramming into the cross-currents present in the gap.
That and cfm is volume not air speed. Also it looked like a lot of smoke was coming out between the fan and the duct.
Lots of static pressure. the fan was built for volume, not static pressure. And the noose was too short and didn't drop air inside the tunnel to draw it back in.
But the neck is too small and thus turbulent. Vanes might have helped.
@@glenmcgillivray4707 True, it should've been a compressor fan, to push air through the nosle
@@kristmadsen What Is Islam?
Islam is not just another religion.
It is the same message preached by Moses, Jesus and Abraham.
Islam literally means ‘submission to God’ and it teaches us to have a direct relationship with God.
It reminds us that since God created us, no one should be worshipped except God alone.
It also teaches that God is nothing like a human being or like anything that we can imagine.
The concept of God is summarized in the Quran as:
{ “Say, He is God, the One. God, the Absolute. He does not give birth, nor was He born, and there is nothing like Him.”} (Quran 112:1-4)
Becoming a Muslim is not turning your back to Jesus.
Rather it’s going back to the original teachings of Jesus and obeying him
@@kristmadsen CFM in a known volume is literally air speed.
The Dragon’s Breathe fan was designed for flow when it needed to be designed for static pressure and allow the flow to come from Bernoulli’s effect. It might perform a lot better if the restriction had a wider outlet hole to reduce the amount of pressure the fan is trying to fight against.
Or, you know, not have a restriction at all, and attach the fan directly.
Makes no sense to attach any restriction, any restriction will make the fan more less air, and for this test you want it to move as much air as possible.
@@jort93z That's ... not how bernoulli's effect work, Think of it as tunnelling air by creating difference in pressure, as the comments said if the tunnel was wider, it would have performed significantly better.
@@amineabdz Tunneling the air is not the issue, you can get that to work, the issue is moving more air than you started with. They want to get more airflow with this device, that part won't work.
Any flow restriction, doesn't matter if it is 10% or 90%, will decrease the amount of airflow. If you want this to somehow get more airflow, you need the area over which the air is spread to be considerably larger than the fan.
The measurement tunnel would need to be wider, and in that case the measurement isn't comparable anymore.
just attaching it directly will be more effective, no matter how they redesign their "Air maximizer".
@@jort93z you still don't really understand this, volume stream is velocity*area, if he made that restrictive tubes diameter about 10-15mm smaller than the air tunnel and made it go into that tunnel, he would benefit from that restriction. Air speed would increase by couple % and the pressure difference would suck more air through that 5-7.5mm gap between restrictor and tunnel.
Dude said he studies mechanical engineering, I guess he should go back to drawing board and pay more attention at college
@@mikoajkosma4010 Well, but you are forgetting that a restriction will make the fan work against a higher pressure, making it work less effectively. If you could just make a restriction and increase airflow, surely you'd just decrease it by 99% and get fat gains, but that's not how it works. The fan will need to work harder and move less air.
Well, when experimenting, some experiments go well, others fail.
The problem with making a constricted air duct is that the back pressure from trying to compress the air fights the already low-torque motor. I had this issue when designing Ov3rdrive as well.
But the flow on the batwing was beautiful!
yeah. it is restricting WAY too much. it should have maximum 20% restriction or so if even and the gap to the duct is way too big too. the end of the restriction should be inside the duct or ending directly at the duct with around a cm ring of gap for the other air around to pull in. (the values are picked by my gut and are most likely wrong or inefficient, but should work better than what was in the video)
Correct, my main complaint about Noctua is that my standard dust mesh on my computer is enough to completly hamper the airflow on the noctua fans, even though they are SP fans. I was very very surprised especially because i bought the 3000rpm pro fans from Noctua, they only states they have a sp of 7.6mmH2O where my current fans has 65mmH20 difference was 15C in case temperature. On top T30 also only have a sp of 3.3 which is completly absurdly low...
Yeah, I think the funnel shouldn't decrease in diameter as aggressively as it did.
I wonder if a secondary venturi tube for the outside air would help...
No flow restriction will help it flow more air, that makes no sense.
It will increase the velocity, yes, but it will decrease again once it enters the test-pipe.
@@jort93z the idea is that it creates a low pressure area that is taking air from the gap with it into the duct like a "fanless fan". the problem is just the way it is shaped and that the idea doesn't really work on such low pressures and low airflow.
A low restrictive venturi tube would be more effective. Since it's a fan and not compressed air aiming for a 5-10% increase in airflow may be achievable.
Yeah... I may have been too ambitious with my narrowing... :)
@@Selbitschka but it was a very clever idea. Tweak the design and submit again, I think I've seen some comebacks in the series. These low pressure "fluid grabbers" when tweaked right are very efficient. I use venturi tubes to inject ozone in my house water. They suck hard (pun intended).
@@Selbitschka Try measuring the pressure produced by the fan. That will tell you how much you can narrow it. Use a length of vinyl tubing with water in it to measure pressure.
@@bladedpenguin Thank you. I am in the process of redesigning it now.
@@RoboticParanoia Working on that now and hopefully can get a comeback. :)
Hey man thanks so much for showing off my fan (Blowsie, Blowie, whatever)! I can't believe it did almost as well as the stock design. I designed it to suck air into the center so I was very happy when the 2nd smoke shot showed exactly that. The low RPM is very interesting, I honestly have no idea how thick the blades are so I can probably lose a lot of weight and get some better performance if I load the design into some beefier 3d modeling software. Anyway thanks again!
sorry i cant read, i called it both blowie and blowsie
Nice work! 😃👍
@@MajorHardware haha it's all good, I put very little thought into the name.
My first fan design performed better than I thought. CHEETAH 👌🏻
It worked like intended, Congratz on reaching top 10 on the fan show down. Impressive.
visually i liked your design. it also looks like it would be a good static pressure fan. well done.
Looks like you might need to make a thin foam gasket or use some sealant. Smoke test reveals where the air shouldn't be flowing quite clearly.
I've been 3d printing gaskets :) Works good once you get the hang of it. So far I've put them on >> rainbird anti siphon valves ; Water hoses ; Lots of water pumps ; Sinks/faucets . I got tired of buying them. They break so often it seems like a conspiracy ! For high pressure stuff, I will include a space to fill with aquarium silicone or E600. PETG
I think it has more to do with the static pressure and that's why it goes out the side, if the pressure was lower the fan would pass the air right by the air gap and out the back like it should. But since the end of the tube is so small the air isn't compressing because the fan is weak, so it just increases pressure and flows out the sides.
Yep there is a lot of leakage....
in the rc airplane world when we use edf motors, there's something called FSA fan swept area(physical area swept by the blades minus the center hub area). how you neck down the tube behind the edf affects your top speed\ max thrust, a thrust tube that is 100% of the fsa will have higher static thrust(torque) and one that is necked down to 90%fsa will have higher top speed(horse power) after a certain point when constraining the thrust tube it becomes inefficient, and unable to push past the squeeze. the max we typically use is 80%. that Bernoulli tube looks like its 50% at best. id put money that the multiplier would work better if Different Fsa diameters were tried.
if the fan is 120mm and the hub is about 40mm then that means 100% fsa would be an 80mm exhaust tube, and 80% of that 64mm
so the workable range from 110% fsa down to 80% would be about 90mm to 65mm.
that JUST the tube being efficient at its job at speeding up the air. this doesnt take into account the distance and area available for the that sped up air to "drag" more air along with it(distance from end of the tube to the beginning of the next "orifice"
id start with an 80mm tube and see what it does, then id go right down to 65mm and see if its any better. it may still not work, the motor may not have enough ooompf to speed the air up significantly. only then maybe start adjusting the distance from the end of the thrust tube.
Your FSA calculation is wrong, you can't do 120-40=80 , you need to calculate the areas before subtracting, the 100%FSA exhaust tube is 113.1mm, 80%fsa is 90.5mm
@@Henning_S.
Thanks for the corrections
Huh this is something i did not understand correctly even though ive built thrust tubes in the past.
I have some edfs lying around should build an aircraft
You should really look up the video of the cheater on here. It has smoothed out air flow behind it because it has like rear cone pod behind the motor. The cheetah fan on that cheater frame would have the best chance of beating it.
The ideal angle for a duct like that is around 11° of divergence/convergence.
Just needed to be more gradual.
Also having the exit closer to the wind tunnel exit would be better.
The dragons breath needs a second slightly tunnel to funnel that air and any extra air the fast moving air pulls in. The problem is that with all the extra open space now suddenly the air runs into turbulence and rapidly slows down and the pressure increases blocking airflow
It's kind of ironic that the idea was to reduce the area but then overlooked the sudden increase from a finite area to an "infinite" area at the end (which to be fair for liquids this wouldn't matter, but this is air which is compressible which always threw me off when I took fluid mechanics). I'd be more curious on a test where this design was mirrored so that it would be a Venturi tube instead of half of one, just to see if it does anything.
I bought Anemometer and made a test rig, I got Bernoulli's principal to work but I used a blower style fan design. The best I got was 660ft per min after 7 rotor designs.
Hey @TheRattleSnake3145 did you try any of the principles @RichardInTN (top comment) recommend like adding a shroud so the air is being sucked in, or changing the fan to be a compressor/static pressure style? Also I like the idea of creating the vanes on the rear duct, no Idea if creating a vortex would increase flow though.
@@mavric1177 I didn't use a shroud for the blower design but when I gave up on the blower/Bernoulli design I designed a more traditional design. The shroud infront of the fan made a significant improvement to flow. My current best design beats the noctua and I believe just beats the cheater. However my actual individual results will be different to his as my test rig is different. My first try at veins behind the fan reduced the airflow.
@@mavric1177 the Bernoulli design just wasn't designed correctly to work.
Venturi ejectors are like a class 3 lever for airflow. They convert high pressure, high speed, low mass flow into slow, high mass flow (pressure is torque and speed is MAF in this analogy). Axial fans are already slow, low pressure, high mass movers so it's like trying to start a bike in high gear.
7:12 Look like the fan is going from low back pressure to high back pressure. Opposite of what you want.
Looks like the dragons breath needs a seal. Huge loss around the Crack between the fan and the funnel. He'll even some tape.
That's the assumption a had myself also if it has streaks inside it would make more acceleration i believe , but overall great Idea , the implementation need some more work 💗
Just too steep a funnel for the end diameter, the air is being bounced back out before it can squeeze down that narrow throat. The spacer is also too large, the end of the funnel needs to already be inside the wind tunnel or boxed in.
If I was to refine the design I'd drop the spacer so it's only as tall as the funnel and wall it in so it only has 2-3 cm of gap where it meets the fan. The throat of the funnel would be ~20% larger with a straight taper and no neck, maybe do some testing to see if vortex generation has a beneficial effect but I'd probably keep it smooth for simplicity's sake.
I was thinking a rubber band would seal it fine, easier than tape.
That "classic" spinning Batwing transition had me cheer out loud! Stellar editing
It's so veiny!
For the dragon’s breath to have a chance, make a good seal between the funnel and the fan
Will that do it?
@@Tgspartnership it will give it an extra chance. In the smoke test I saw a lot of air escape through the sides where the funnel connects with the fan.
I'd like to see the anemometer placed right at the end of the cone shaped thingy. I think a big part of the problem is that not only is it constricting the flow of air going through it, but as soon as the air gets to the end of the cone, It just dissipates out in every direction because it's no longer constricted. But the air coming right out of it might have some speed
I think if the constricting tube was halfway placed inside the air tunnel it would help out a bit. (Plus a bit o' size increase of the end of the funnel could likely help)
You should try more testing with the t30. It would be cool to see fans that would be more efficient with more rpm/ torque. Fans like the contra rotating one and the jet engine compressor style fan from about a year ago.
Also I’ve seen this a lot in your yt comments. But try making different fan classes. One would be a standard class that it basically all the fans that fit in the frame of the a12x25 and ones that have external components outside the frame of the a12x25. The external class would be interesting with fans that don’t work out like the dragons breath.
Love the air pattern out from the bat fan! Looks really straight and great!
I've noticed with all of the air flow modifiers like this one, the Dyson style, etc., that they would work if the motor fan produced a significant amount of power like a vacuum motor. The challenge still remains to create a low power fan to increase it's airflow as what people have been trying. I think if this dragon's breath adapter has the same outlet diameter as input, then introduce the venturi effect they are going for, it may work! Or make the outlet diameter size as if the fan hub was not present. So just blade diameter with no hub. Just a couple of ideas. Love the series!
i think the cone would have a better chance if the smaller end is inside the wind tunnel, with only an inch or so of air gap to suck in air. maybe even better to add fins on the outside of the cone to align outside air coming in.
Bernoulli requires pressure potential to be exchanged for kinetic potential. Fans accelerate air flow, they do not pressurize air. This gets into a deep engineering field of fluid movers; pumps, compressors, and fans. In compressible fluids movers are separated into fans and compressors because they add energy to the air in fundamentally different ways. Fans do not compress air, they simply add energy to the air in an incompressible fashion, similar to incompressible fluid pumps. Compressors on the other hand are moving with such speeds that the fluid can accumulate pressure potential locally without necessarily accelerating (it happens both ways) meaning the fluid can come out of a compressor with the same velocity it entered but at higher pressure. With a fan you are never going to get that compressible behavior because they simple do not move fast enough. The whole point of fans is to add small amounts of energy to a lot of air. So back to this design, if fans simply increase the velocity of the air, then any area reduction will further reduce the pressure, and now we run into a conflict. Fluid will flow from high to low pressure, and if all we did was increase the velocity then we are now below atmospheric pressure. Meaning are total flow must reduce to keep fluid flowing in the correct direction. The whole goal of this exercise is to design fans that maximize the energy exchange efficiency between the motor and the air. This means low turbulence, one dimensional airflows along the axis of the fan. Just some tips if you are going to investigate along this path, constant flow areas and flow stabilizers are going to give you the optimal performance
This continues to be one of the most entertaining interactive ongoing competitions on UA-cam. Than you so much for keeping this going.
I would try to put the fan on the opposite end of the flow maximizer and redo the smoke test. That would be more inline with the Bernoulli Principle.
yea I don't see how this would work, it just creates a high pressure at the end of the nozzle and it goes everywhere from there
I'm not exactly an expert but I thought he had the airflow backwards! Pretty sure the idea of speeding up the air coming in would be an excellent way of making a better fan.
Woohoo, new Major Hardware vid!
Looking forward to seeing the new fan mount in action. Always an awesome evening view when a new Fan Showdown episode drops.
Thanks for the awesome content and constant improvements!
I think the best part about this video is how he doesn't even realize the pun when he repeatedly said he's a "BIG FAN" of the new purple print material.
With the "Dragons Breath", maybe a suggestion is to shorten the "legs" on the funnel shroud, so the funnel is ending inside the wind tunnel? I am not sure of the physics, but that feels like it might be more effective.
The problem seems to be the amount of pressure it has to overcome to push the air through the funnel, not so much the funnel end
If you look at the outflow of all the fans, most of them produce a significant amount of radial outflow. If the dragon's breath also had this issue, then the radial outflow from the fan would hit the channel wall almost perpendicular, leading to more collisions with the axial outflow and thus a significant decrease in exiting flow velocity and a resulting increase in pressure, meaning it had to fight even more static pressure within the funnel.
A channel converting this radial outflow to be more axial could have a better effect than just a normal isentropic nozzle
so glad that you added the wall for the smoke test, it makes the visual a lot easier to read and understand what is actually happening to airflow.
The size change of the funnel was probably too much too quickly, so the air would rather blow out the sides at the front than go through the funnel.
A less severe funnel might work.
I'm super glad I started watching this series since either near the end of S1 or the beginning of S2; always satisfying to watch the smoke flow through the fans and the rad 80s-style music you picked starting from IIRC early S3 (thank you again for identifying it as a Mattie Maguire stock music track) has been a great choice for the showdown montage! :)
Wow! Over 200k subscribers! Congrats Major Hardware, your hard work, consistency, persistence, and viewer engagement is a model for creating a community on UA-cam. Here's to the next 100k!
I feel like Chad did a lot of fluid dynamics math and just completely ignored the fact that CFM through a fan is totally different than CFM through an air pump. This is a really good object lesson in what backpressure is and how it works (and in the case of an axial fan, how it makes things *not* work).
@ Major Hardware add some some tiny holes on the funnel part of the dragons breath to help alleviate some of the drag inside like in the engines of the sr-71 blackbird
I stopped the video at 6:50 and literally exclaimed aloud, "that's a good fan!" I think that bat one is going to surprise us
Your videos are excellent at displaying the airflow capabilities of different fans.
This is why I'd like to request a video demonstrating different CASE airflows.
Idk throw in a cardboard blockout to represent the general shape of a gpu and ram, for the air to swirl around. Naturally, I'd like to see the most air possible dragged across the gpu.
--Then maybe compare fan positionings for airflow. 3 fans vs 2 fans, 120 vs 140. Using a vent for either intake or exhaust. Or just displaying how different cases, in general, respond to the same setup.
--Like, do their more-or-less-the-same dust filters change the airflow in the brands' case?
--do the cases that only have vents on the sides of the front panel, surprisingly, actually pull the same amount of air?
--does the latter change how the air is flowing internally, because of the angle the air is being dragged in?
Fan designers need to keep something in mind: Virtually every fan design does not merely push air in one direction. It also pushes air outward.
When that air moving outward hits a sloped wall, it is "reflected" by that wall. A straight wall will have little effect apart from some friction, because while the air might "bounce back," it's still moving in the direction of flow at roughly the same speed.
If the funnel constricts, the air will "reflect" back toward the fan. If it widens, it will "reflect" in the direction of flow.
Nobody, from what I've seen, has taken advantage of the latter, and many have fallen prey to the former.
there really should be a catagory for fan add-on prints that just get testing with a stock a12 as well as whatever the creator sends in. i'd love to see how this and the dyson perform with a standard fan. that would actually be useful for someone who doesn't want to buy all new fans but wants better airflow and has a printer and doesn't mind a unique style flair
Idea for the next season, use the T30 as base model, this should allow much more wild and big design as it have so much more torque to play with.
The main issue with the dragons breath is vortex that develop in the funnel, I ran into this same issue when trying to make adapters for 40mm noctua fans for my 3d printer. You can see the issue in the smoke test where air is pushed back out the fan.
You can mitigate the vortex by adding stator veins in the compression chamber, but over all you will lose output not gain. (the part cooling adapters i made did not produce enough static pressure after adding those veins to cool PLA so i gave up on the idea and went back to using dual 5015 blowers)
(the info i learned this from was some HVAC documentation I randomly found so yeah apparently this is a known thing in certain circles)
to cool pla with 40mm's you can't(or don't much need anyway) constrict the flow, instead an open air guide works well enough. needs insulating the heatblock tho.
that being said I have one like 10watt 40mm delta I can swap if I need more cooling.. that thing blows. downside is that it sounds like a vacuum cleaner.
@@lasskinn474 Yeah i was going for a silent printer :P
As for constrict vs open, depends I guess? I was aiming more for a general purpose design as I print other materials as well. Either way the experiment wasn't worth over using 2 good 5015's.
@@nobodytoyou4887 yea i just cut power to the fan when printing abs, but a constricted highly localized constricted one would be more useful with abs I guess if printing small bits, so as to not cool the entire piece but not have it mushy up small peaks.
There seems to be a lot of air coming out from the sides. Maybe using the "radiator gasket" or whatever Noctua calls the gasket that goes all around the fan to seal it to radiators could be a good idea for testing in general? Less losses!
It's quite clear that there can't be any openings between the fan and the air intake, or else most of the air will just escape from the openings. The idea of a longer tunnel could maybe work better if it's completely sealed. Perhaps it could have two fans, one on each end, or something.
I really do love this series! And I'm a biochem engineer who's never built a computer in my life....you just make it very interesting, the good old trial and error engineering really speaks to me
Biochem, eh? I think you know what needs to be done...
_Design a fan that looks like adenosine triphosphate._
@@pirojfmifhghek566 haha, yeah that won't be difficult to print at all! Also, if the tips hit the casing they'll turn into ADP which will deliver far less energy ;)
It's almost as if conservation of energy is actually real.
Perhaps the 'Dragons Breath' would have worked better with the wind tunnel if the 'flow maximiser' had a larger tube diameter and lesser gradient, and for that tube to protrude into the Wind Tunnel for say 10 cm, with a clearance of say 1 or 2 cm between the inside of the Wind Tunnel and the outside of the 'maximiser' tube.
This should allow for the lower pressure to be generated at the wind tunnel opening. I guess I'm saying like an inside out Dyson fan.
Sorry I don't CAD, so I can't draw it.
Should try a rifled tunnel and a gasket between fan and tunnel to prevent side leaks :D
Bernoulli and laminar flow seem to go hand in hand, im not sure if that design managed it there but it was still a brill design and concept, honestly a good fail can really beat a poor win. I signed up, great series, looking forward to seeing more contendors 🙂
Watching today's episode just seemed to reinforce my idea that the smoke needs to be more uniform.
What about a nozzle that attaches directly to the smoke machine and creates a curtain of smoke instead of a linear flow, that can then cascade down in front of the fan to provide a more even flow that demonstrates the fans' airflow with complete independence from how you squeeze the bottle, or where it pours out?
Fans just need attaching to the other side of the plexi glass.
4:43 Cheater: "You're just presenting me with more parts to make me even stronger."
Here's an experiment I hope you'll do: The bernoulli fan's funnel-and-supports, and make the supports half the height, so the funnel sticks a bit inside the wind tunnel. Alternatively, extend the funnel with a pipe part which is half the height of the supports. Getting the smaller pipe inside the bigger pipe would increase the chances of the funneled air to pull along the outside air, because the current configuration makes high pressure air which then spreads in all directions instead of continuing straight through the wind tunnel. You could also test those changes with fans with a higher angle of attack, to get more static pressure at the price of less airflow, to deal with the loss in efficiency caused by the fan needing to compress the air so it fits into the funnel.
>Be Chad.
>Walks in.
>Yells with an open chest: "Cheater beater."
>Fails even in being registered by the sensors in testing.
>Refuses to elaborate further.
>Leaves.
What a Chad.
I think that a resin 3D printer would be much better for fans. Less air drag from imperfections
10:34 would be very helpful if you could also measure noise of T30 🙏🤞
since you did for the other reference, Noctua one. thanks!
2:42 - One of the best unintentional puns in history.
_B I G F A N_
You should try it again without that massive plastic frame structure!
Let the bottleneck portion enter into the tunnel rather than let that plastic stand set it an extra foot away -letting all the ambient pressure out.
The bowl of the funnel from that multi-part fan was the problem. The bowl didn't have enough depth to it before it got smaller. That created a buildup of air that then slowed/stopped new air from being able to enter.
Gotta love how all my favourite channels have been uploading while I'm dealing with covid hehe. Keep up the great content!
S5 should use a Delta fan as the hosting unit for all the blade/propeller designs! They are insanely string units usually used for industrial server applications.
It would be cool to see an all-out, as wild as possible, competition!
Would love to see a bonus episode where you take some of these fans and pair them up with other ducts people have designed. Maybe even specifically testing other fans with the cheater duct?
E4 Mafia!
The thing about the Bernoulli Multiplier is that the shape was meant to take a high pressure air stream and convert it into a high volume air stream. If you already have a high volume air stream it won't do anything.
The shape does work on emergency escape slides.
I'm so happy for a non-working idea. Those can turn out to be the way to learn. And I've already seen some comments why it didn't work and how it could be improved. I hope to see the continuation of this design.
This is my favorite color from all the ones you used! Damn it is beautiful.
9:10 idea was sound but he reversed the tunel, air is fastest near blade edge, so if you would do that and make an outside ring that is the push (like dyson is doing) then you can pull much more.
problem is that this depends on PRESSURE and you seen how easy it is to leak some.
I am sure it would be much faster if fan would be duct taped to the tunnel, stopping the precious pressure buildup from leaking everywhere...
I think you almost need to mount the deagon's breath backwards so the fan pulls higher pressure air from the immediate front of the cone, allows it to expand in the fan and go out the back
With standard bladed fan designs, venturi tubes don't work on output, only input, they simply do not have the pressure capabilities to push the air through a restriction and it will just bleed out the holes between if not sealed well enough. But, if you truly want to use a venturi tube, you have to put the fan half way between the intake curve and the output hole and seal the fan housing properly.
The plexiglass holder is a great idea. No more bleed over by the smoke going around the fan.
Bernoulli's principle also states that with a decrease in static pressure comes a rapid decrease in temperature. If you give that air room to expand before it hits your heat sink then you effectively have the standard cooling from the air flow over the heatsink, but also some adiabatic cooling from the rapid expansion of air
You could potentially improve the Dragon's Breath adding ANOTHER tapered mouth on the sucky part of the fan. You would also need to seal the edges otherwise that pressure is just going to escape.
Just for grins... Take a couple of cuts at the maximizer... Start wit an inch, then 2, then 3.
I think what the problem we're seeing is too much pressure drop in the tube in its attempt to speed up the air to drag more in with it.. By cutting off progressively more, you'll reduce pressure and get more airflow through it which should gain some numbers.
Bernoulli's theorem just says that you're basically concentrating the pressure into a smaller volume, but it adds drag so you'll only ever lose pressure from it
I need to catch up on this series. That Dragon capsule is sweet!
Thank you so much for featuring my Dragon's Breath fan design. I am in the process of redesigning it now and am actually going to do some testing to see if I can get the desired effect I was going for. I hope that the redesign can be given the chance to redeem itself. :)
Hey,grats on your solo fan design. Is there a chance you could share the files for your fan for home usage?
One of the difficulties with putting a converging nozzle on a fan producing substantial pressure is that the output flow (from the blades) is (usually) quite rotational. The bits of air move in a helix.
Reducing the diameter of a rotational flow means squeezing the rotary motion down to a smaller diameter, in a direction anti-parallel to the centrifugal force of the air. This work, basically the opposite of what happens in a centrifugal blower, produces a force that opposes the forward flow of air.
I am not sure how big this effect is, or how hard it is to remove the rotation from the flow with a set of stator blades or similar.
Suggestion for the fan supporting wall: Either use some tape around the fan frame's perimeter to keep air from leaking through the gap between fan and plexiglass, or get a watercooling radiator fan gasket so it seals correctly. If you look carefully, during the side-on views you can see smoke sneaking out from between the fan and the wall.
Its like putting your thumb over the hose and the water shoots out faster. Just because its faster doesnt mean that more air is flowing
I dropped out of this since S2 or something, but wow, I'm amazed to see it's gone on to season 4 now. People are really keen on their fan design skills haha
_Also the batwing smashing the smoke test 👌🏻👌🏻_
One main factor for failing is the imperfections on surface that increase the drag and consequently create turbulence and loss of airflow
What a missed opportunity to print the first one rainbow-colored and call it David Blowie.
The problem with the bernouli fan is that the frame was designed to be surrounded by the wind tunnel not open air. The concept is that the air in the tube already creates the pressure and the air coming from the fan is accelerated by it.
I would love to see a test that shows resistance to recirculation. An example of what I mean are some mid-cooler fans by Zalman: they initially shipped with plastic rings around the outside, but overclockers discovered they are more effective with the rings removed. My rudimentary testing with little strings stuck to the interior of my tower suggested this is because the air spilling out the side of open fan blades creates a barrier that reduces recirculation of warmer exhaust air back around the fan to mix with the cool inlet air.
i think there is clearly leaking pressure of the dragon breath after seeing the smoke test, prob it wouldnt imrpove much but there is a gap between the noctua and the air duct on atleast 2 sides
The end of the adapter needs a bell exhaust for it to work. Also the narrowest portion of the throat is too close to the fan intake. Small changes should big a very big difference in the end test.
I think a lot of designers seem to underestimate how impactful back-pressure is. If you have too much back-pressure it creates a turbulent flow before even passing the vanes
The cone needs a gradual longer change in size to allow air to funnel and not abort as it gets compressed since it’s an open air design. Or you can have a longer large volume of high pressure low velocity necked down to generate low pressure high velocity (which your design has) by calculating the size of tube needed to optimize your output with the fan size (Venturi effect). Just my $0.02
It's funny, if you put the flow maximizer on in inlet instead of the outlet, it would at least increase pressure but in it's intended use it appears to be a flow minimizer. lol
Consider mounting the fans on the back side of the partition so the inlet is flush with the wall and the fan thickness doesn't disrupt the fog flow.
Big fan of the new filament... Lol.. That unintended pun was excellent.. :)
One problem with the "flow maximizer" is that it makes the tube much longer than needed. Also, the smoke test shows that the air comes out of the back in a 45° cone, which can't work with that gap.
If it was only leaving half an inch or less between the fan housing and the wind tunnel edge, it would at least spit out its air partway down the wind tunnel, where everything COULD be redirected forwards.
love the plexi fan setup as it separates sides…that dragons breath is thinking backwards. he was try to make a velocity stack, those work on suction. that fan isn’t the right type to create static pressure or act as a compressor…
I think if you want the Bernoulli tunnel to work you need to run the end through the mount. That way you are pulling air in the tube. This one has too much venting.
The bat wing plus was my favourite among these.
I am surprised how well it performed in the smoke test
the tune at 5:34 always reminds me of an abroad in japan video.
Also is needs to be "seald" now the air can basicly go out the front of the "mesuring tunnel"(?) Or not in to it much at all.
I'm watching smoke pour out between the fan and the "duct". You have to at the very least make the connection airtight to give it a fighting chance. Any air pressure being built up just flows right out the non sealed joint.
If you were to put some duct onto the front of the fan it would create an area of low pressure which would help drive air through the nozzle. Referring to the constrictor nozzle thing.
I'm at 5:03, haven't gotten to the tests yet. I'm guessing the 'cheater beater' will fail. You don't get any more air volume through, compression of the air by the reducing cavity increases flow rate at the smaller diameter. But if anything, your going to reduce overall air volume through the system by the resistance of that compression.
If intake nozzle is smoother and rounded it can really help with Bernoulli principle
Holy FRICK. This particular video was a mind blown moment.... the reason why turbine inspired fans don't work as well is because turbines have to compress the gases. They're not just moving air fast, they're compressing it. So the fans without housing/flow modifiers are going to do better if they have some kind of progressive pitch/lift generating shape with low to medium blade spacing rather than flat pitch high blade count.
You can buy 120mm rubber gaskets for the fans on amazon for cheap, removing the corner rubber is necessary for a tight seal. I believe you will see a increase in static pressure when eliminating the gaps. I did this for 5 radiator fans and dropped my average water temp
Dragon's Breath expectations: Feels like that's just going to make a lot of unnecessary resistance, then turbulent air escape before it enters the wind tunnel.