Yeah those IGVs look a bit much, both in number and trailing edge pitch.

I agree with this. It is taking laminar flow input air and making it slightly turbulent. This reduces the efficiency of the compressor, generating a slight stall effect, in addition to the back-pressure generated when the front of the compressor blades sweep across the back of the static input vanes.

Plus the acoustic sheer is dissipating fan energy if that is the close tolerance on the front

Plus the #s are so close errors in humidity, temp and power grid voltage could be different considering how long ago the cheater was tested to be comparable so a back to back rematch is needed. Get on it ! Bill Burr acting as Steve jobs

I don't know the science behind it, but I think this would make for a good test. Can you please try it?

This is a pressure issue. If you notice on the angled shots during the smoke tests, the hole in the plastic doesn't match the duct of the A12x25, this block in the path will cause the flow to not only change direction but create a pocket of pressure; this also happened with every fan shown in this episode, and multiple fans in past episodes. If you also notice during other tests, the amount of leakage isn't consistent, which is more than likely the plastic warping under pressure. Gasketing of any kind won't actually help this, a more rigid material with a better cut hole would be a better start, as this would be taking care of the primary issue and not putting a bandaid over it. Gasketing, which is also needed in these tests, only fixes poor tolerances between surfaces, for example how the fan's anti-vibration corners add a gap between the housing and mounting surface. The actual testing tube is much better off, as it is better designed. Gasketing isn't entirely needed for high airflow, but is needed for high pressure, as high density fins such as those in radiators need higher pressure to push air through smaller gaps, and this is where pressure-based leaks cause issues, as you'll see a loss in pressure when the leak occurs. As for gasketing materials for tests, there's really only one optimal option, a thin sheet of closed cell foam should suffice, or a similar crushing surface. Putty is too permanent for testing purposes, plus there's no guarantee of a good surface unless you're finishing a part against a known flat surface. For all intents and purposes an O-ring would be too difficult to incorporate on these for testing purposes. Another option is using tape, but that's more of a bandaid as a pressure leak will also balloon any excess slack in the tape which then affects the dynamics of the airflow within the system, and if the tape fails under pressure then you have more issues. Thus why a crushing layer is more optimal, but then this also runs into the issue of the smoke test setup, quality of manufacturing the part. A lot of people on this channel will also claim that none of this matters, as this is all for entertainment purposes, even though we're being shown somewhat scientific testing. I also think it's funny how this supposed non-issue gets pointed out more and more on each new video; I guess that supposed non-issue actually is becoming a noticeable issue among the audience. Sometimes I wish GN had a resin printer or two on hand to do this kind of testing on actual test rigs used in the industry, plus they're pretty decent at having good test methodology, not only would the results be better, but you also don't get idiots white knighting the creator over poor practice. But as far as the issues actually getting fixed, I don't think it'll happen for season 4, doubt it'll happen for season 5 either.

@@xaytana Just add sillicone problem solved. Hehe many design oversights are almost always in assembly process. Haha

@@alesksander Lol!

I bet the extra 11 CFM was lost to the air leaking it the sides...? I propose a rematch!

@@xaytana solid observations. Most of us do view The Fan Showdown for it's entertainment factor. I personally feel like the creator has, over the seasons, significantly improve on building and testing techniques. For what is worth, The Fan Showdown started out as something fun and had grown to be something quite a bit more interesting than I anticipated. I tune in solely on the hopes of seeing The Cheater and finding out how. O so fascinating 😁

Let's make this happen! I want to see the cheater fall :D

He could easily print a tpu gasket.

Rear stack needs to be redesigned, it's too restrictive, that's why you're seeing leakage, because it's generating too much pressure. Sealing it with a gasket won't solve the problem really, might do a little better, but, you're leaving performance on the table doing that.

Also at 5:36 you can see that one of the blades is not perfect. There is room to improve.

thank you guys, cheater has the advantage because his blades are printed in different orientation so they printed better. Now I know at what rpm the fan spins so I could do a bit of tweaking.

I was thinking that too

Indeed

FINALLY SOMEONE SAID IT GODDAMM

I'm honestly shocked and disappointed it isn't called the "Gratest Fan"

"The Greatest Fan" I'm sure that should be "The Gratest Fan"

Maybe tweaking an existing design already high up on the scoreboard would be ideal. What about bumps on the blades and intake lip? For example, whales have evolved "tubercles" on their fins to help with laminar flow. Noctua uses something similar on their fans, both for performance and noise. More on this: en.wikipedia.org/wiki/Tubercle_effect Not sure how much this would help with such rough 3d print surfaces. Tubercles need to be smooth for proper flow.

yes that's why i watch this series actually, in hopes of somebody finding a blade design i can stick on my 3D printer's part cooling fan

yeah, maybe with proper balancing of the blades, smoother surface, less vibration. All these small changes together could make a big impact

Damn, your comment was on top and spoiled it for me.

yessss

Cheater has a small edge in airflow, but is leading by a significant margin in noise.

thank you, I honestly expected worse, so this is a pleasant surprise, an I knew that it is going to be noisy, thats why the fan has 8 blades and the intake and the exaust have 7. This gives a higher frequency to the fan , but I was hopeing it would decrease the volume.

@@frandiminic3559 Very cool! A few tweaks and you will be on top.

Yeah man, that RAH-66 fan impressed the crap out of me. Same goes for the squirrel cage/washing machine design.

Then you should also retest the cheater and every other fan.

It could be calculated based on the amount of power the motor motor uses at full speed. The amount of energy being imparted to the air must be equal to the energy input of the motor, assuming an ideal motor and fan. The energy being added to the air is proportional to the increase in velocity, the increase in pressure, increase in heat, sound energy released, and mass of air being moved. For an ideal fan, the heat term is removed and the sound term is negligible. Working through all that math, you can isolate a velocity as a function of open area of the fan exit (all other factors collapsed into a constant). Plug in the fan exit area, get your velocity, then multiply those numbers together for air flow rate.

@@tronique5736 So easy! Maybe for you. My mind is hanging up on questions like whether even distribution of energy across the entire fan exit area is possible, degree of air friction losses, number of atoms of air in a given volume at a given pressure and temperature combo, possible effects of pressure reduction at the face of the fan on the volume of air that can be moved, reduction in effective volume of air moved as a result of swirling flow... so many things to consider, so many variables, so many things to be balanced... Can't you just spit out a number for me? Something like "1200cfm at sea level, 70 degrees, 10% humidity, zero ambient air flow, friction losses ignored, stater vanes to optimize air flow, no effects from turbulence, consistent flow across the entire face of the fan, no back-pressure..."?

@@warpedweirdo That's basically what the calculation I outlined would do. I have no idea what the actual number is, but that puts an uper bound. Then you start trying to model unavoidable losses to get closer to a more realistic value. This process isn't easy at all.

Reminds me of the old analog film camera reels

damn nice

Congratulations 🎉 on your wedding!

I believe he did so in episode 6 or 7

@@jawms oh good stuff i'll have to go back and review though episodes

most definitely. It should've been called the gratest

@@boriscareaway5862 gottem

"Give me the happy chemicals 3D-fan boy"

Umm. The cheater has been number 1 for like 3 seasons.

@@Temperans The Cheater was the winner of Season 3..... The Cheater has also never gone up against the winner of Season 1.

this is a 100% UNDERRATED COMMENT.

Yup. Definitely looks like a grater!

my (our?) first thought!

I'm leaning towards a film reel...

The file is literally called 'gratest fan' so yeah. He totally missed the pun.

new fan showdown idea, but instead you print front top and rear pannels on something like a diy "modular" case

Little companies want to spend engineering time on cases unfortunately

I could print one myself and put it on a hdd motor, a stock shitty fan runs on about 5000-6000rpm

I kinda-sort-of-agree.. but.. the "noise" of fans can be vastly different (even in a comparable db-range) and for some people the "loudest fan" is "more quit" than other "quieter" fans while for others it's the "loudest". I would be interesting, but loads of now defunct "overclocking/computer-websites" have tried it and the tests are usually useless - because the noise-limits is set "incorrectly" (what is correct or not, is not very apparent - personal preferences etc). Noise-limit tests: a pit of despair. We may learn something, but after our host have got three strokes and two heart attacks, by comments/content-creators not understanding the difficulty or have any way to convey the "numbers" (except dB/cfm) which is 100K dollar setup to do an actual "useful" measurement. When the tested fan is used elsewhere those numbers don't really mean much. for end-users, all the numbers are "useless" (there is some use, but it's a difficult topic, without writing a book) when a fan is put on/in different heatsinks/chassis the fins of the heatsink or the chassis may "harmonize" with the fan an create vastly different sound-scape. Then your left with a fan that is "quite" with some irritable characteristics ( very setup-dependant, but very noticeable ), but doesn't really move much air at all. so you buy another one, then another one - finally you just look at static-pressure (temp tests on a rad/heatsink with a dB-level that you can live with) and go from there. Months wasted on useless numbers instead of understanding that if you want to move air: noise is created. Less noise, less work (very simplified). TLDR: Standardized "noise-limit/ceiling" tests are useless for many reasons, their "only" use( again, it's complex topic) : just another test of dB, that you can compare with other testers/websites etc. Usually it ends up with you buying a fan, which sucks(or doesn't, pun intended or not intended..I'm confused..) in everyway other than the amount of noise it makes. Afterwards you end up buying a different fan, that actually moves air and you live with the results. The results: Your temps are okay and the fans are louder than other irritable noises (inherent with computers: buzzing, whirring, clicking, whining). Endnote: Be careful with focusing so much on sound, because there's always one more irritable sound that becomes apparent, when you try to reach lower and lower on the dB-scale. You can and will become "crazy". Take care :)

or a single laser with a spinning mirror

@@insu_na oh look, Fizeau is back at it again with the spinning mirror lasers

I have 4 sharks with lasers on thier heads.

@@insu_na the mirrors your referring to dont spin at all, they vibrate insanely fast, I literally make the motors that hold the mirrors for a living.

@@FhangMedia I literally proposed a DIY solution. I don't care if mirrors in commercial laser levels vibrate instead, but glad you can make a living off of that

What I love is how meta this is. Fans making fans for the fan showdown where the fans face off with other fans using the fans they designed.

it has a cone inside but it is not visible, agree on the x part, but it would unswirl less, nontheless worth trying

Hahha I watched the video and seen the same, that's all on big daddy youtube.

unswirler was made specifically for the speed test, if the test was on the graphic card or any cooler, it would be different design

He tested it in a previous video and the answer is not much

nothing. all about flowrate no pressure. everything for blingbling no oomph on an actual alu finned cooler.