Turning the Sharrow Marine Toroidal Propeller into a PC fan | Fan Showdown S5E6
ฝัง
- เผยแพร่เมื่อ 5 พ.ค. 2023
- Play World of Tanks here: tanks.ly/3nmdN8S
Thank you World of Tanks for sponsoring this video.
During registration use the code COMBAT to get for free:
-7 days premium access
-250k credits
-Cromwell B (tier 6) British premium medium tank
-3 rental tanks for 10 battles each: Tiger 131 (Tier 6) German heavy tank, T78 (Tier 6) American tank destroyer, and Type 64 (Tier 6) Chinese premium light tank
The promo code is only for new players during registration.
I while back I made a video where I tried to take MIT research into toroidal propellers and use it on a PC fan. In that video we spoke about a propeller that's been available for boats for some time. Well today Tom decided to find out what would happen if you took the Sharrow marine toroidal propeller and turned it into a PC fan.
My Twitch
/ majorhardware
Material Used Ziro PLA
amzn.to/3B2Xpx8
Fan Showdown Instructions
• EVERYTHING you need to...
Fan Dims
www.thingiverse.com/thing:455...
Submit your design (at least .STL file)
fanshowdown@gmail.com
Cameras used.
BlackMagic Pocket cinema 6k pro
amzn.to/3m2RLVw
Reddit
/ fanshowdown - วิทยาศาสตร์และเทคโนโลยี
Play World of Tanks here: tanks.ly/3nmdN8S
Thank you World of Tanks for sponsoring this video.
During registration use the code COMBAT to get for free:
-7 days premium access
-250k credits
-Cromwell B (tier 6) British premium medium tank
-3 rental tanks for 10 battles each: Tiger 131 (Tier 6) German heavy tank, T78 (Tier 6) American tank destroyer, and Type 64 (Tier 6) Chinese premium light tank
The promo code is only for new players during registration.
Would love to see the radish get a blank square spacer so the back piece can still get used. It looked like it had a very focused stream in the smoke test. Great work!
Dont Mind World of Tanks, but warthunder is alot better for everything in one game and how they have every tank instead of hitpoints they have models inside
Those resin prints came out looking really cool. Would it be possible to print out a frame as well and transplant only the motor?
It’s a scam…..it’s not actually him…scammers r making waves in the TH-cam comment section nowadays. It’s just sad that people r falling for that. They keep doing it because it works.
I used to play WoT when I still used windows.
The exhaust on "The Radish" is VERY streamlined. His design worked! That's some gorgeous fluid dynamics.
Yeah. I'd love to see that one tested with a spacer so the stator and spike can come in to play.
@@Loebane or the spike cut off. Well, both versions would be best :D
I wonder what the optimal spacer length would be. Might be good to test with multiple lengths so the air has some space to straighten out. Also, that airflow is so smooth I bet it would make a great cheap wind tunnel on a budget.
Major Hardware has some old fan bodies that he could use as a spacer, that would be a easy fix.
@@Loebaneexactly what I was going to say
The results of fluid dynamics engineering are way cooler than the actual fluid dynamics
Theoretically we have all the 3D models, and the experimental results.
It's time to train an AI.
@@jtjames79 This! We really need to train an AI on this data and let it start narrowing designs down around metrics like efficiency, CFM, static pressure, etc. based on preset parameters, like certain types of blade designs, pitch, frequency of blades, etc.
@@AlphaMachina "ok google, make the best pc fan ever"
LOL
@@pvic6959 we are not that far from something like that being possible.
The advantage of the Sharrow propeller is that it has some of the advantages of a ducted propeller without some of the disadvantages. The advantages has already been tested and proven, its just that it is proprietary and expensive.
It is highly applicable in boats, not really applicable in PC fans.
Really love the switch to resin printing. I know it can be hard to work with but from the viewers perspective the higher quality prints it produces is definitely worth it.
So clean!
@@ExclemationMark agreed!
and not to mention more perfect condition
The Elegoo Jupiter also has a build plate big enough for two 120x120mm fan parts side by side, maybe even 4 at a time if you tilt them to the perfect angle. And with resin printing, your print time is the same per unit height, regardless of how much stuff you pack on the plate. So even though washing and curing and dealing with chemicals and ventilation is a pain, it's possible that the resin printing workflow actually ends up being less time-consuming than FDM in the end.
These look like FDM prints, you can see the thicker layer lines in a few places and I've also used this filament before
@@RavenBomb123 I aint no expert. Parts like the grey shroud for that one fan sure. Thats probably FDM. The light blue parts look like resin to me. If its FDM that's even more impressive, though I doubt that.
I'm not an engineer, I'm not a PC uber cooling freak. So why do I find these videos so fascinating? Great stuff from people and great presentation. Keep up the great work.
We never know what were about to see in this series. I love it
I'm not porn star, but find adult entertainment interesting.
The sound of Triskelion was very pleasant. Would love to see more toroidal designs like that so we could get a noise/frequency spectrum comparison!
Some really good entries this round! That Radish smoke test looked pretty good!
Thanks for including the triskelion. The sound profile was actually pretty good as well as the flow test, but not a pressure fan by a long shot. Still better than all other toroidals tested. FWIW, My Nacho fan will blow all the other fans away.
WOAH. The print quality you've achieved here is OUTRAGEOUS. Dang, now I really need to finish that model I was working on.
These look resin printed.
@@JJFX- They are. He mentioned in a recent video that, after comparing old designs on his resin printer he got better results so now every fan will be resin printed where he can manage it.
@@amani576 Makes sense. Will probably be a bit more brittle but I think he just used PLA anyway so overall it should be the better way to go.
yes, indeed!
After years of fdm printing I finally got a resin printer and I’m loving it. The quality is incredible.
The radish was really good, but that turbo pump was very quiet. I think the slightly reduced cfm was well worth the trade off for the super quiet operation!
It almost sounded like the XL was hitting the shroud. Perhaps this could be checked? Also, in terms of testing the Radish, a spacer could be made to give space between the radiator and the flow stabilizer I assume.
Just don't print the cone, its effect is probably really small.
@@ameunier41 It's probably affects the noise profile (and drag,) I think it would work great in a pull config (on front intake) if it was also the body for the motor because now the stock NF-A12 body interferes.
@@submijiru keep in mind I am taking about the scenerio where the fan is clamped to a radiator.
@@ameunier41 Sure, but you still could clamp to a radiator.. in a pull config. And there might be more benefits to keeping the cone, because there'll be more directed, less turbulent airflow and maybe if you do a push-pull with other fans it might have a dampening effect? I don't know, it's worth testing.
The Turbo Pump is notably quiet!
SCAM ALERT !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
He should redo his design and just forget about making it smaller.
The airflow of the Radish looks rediculously insane, if it can sustain this airflow with a rad or towercooler between, this thing looks like a dream if you have your tower setup accordingly. This even looks better than the Cheater. That pulling and push force just seems to be insane.
The advantage of the Sharrow propeller is that it has some of the advantages of a ducted propeller without some of the disadvantages. The advantages has already been tested and proven, its just that it is proprietary and expensive.
It is highly applicable in boats, not really applicable in PC fans. Ducted fans will still be better for air propellers vs. toroidal.
That's why the stock Noctua seems unbeatable. The frame of the fan already acts a duct, no need to overcomplicate things. And it's not like they didn't do their homework when they designed it.
Turbo pump is amazing performance to noise ratio. It actually beats the noctua when you consider that Decibels are logarithmic.
I think you should do better two kinds of fans: the vanilla ones (those who stay inside the normal A12x25 case) and the non vanilla ones (bigger than the normal fan). That way some people might get inspired and create some good designs that can become contenders in the market if something like what happened with the Acceleron.
Edit: HOW MORE THAN 50 LIKES!!???
Edit 2: What... More than 120 likes WHAT!!??
That would honestly be a nice change of pace so everything doesn't come with giant shrouds.
@@Zarincos agreed! longer yploads time and longer duration, with more fans with 2 category
Freestyle and Vanilla
Yoo this is popping up!
Agree with this
I do like your thought process but I don't think the vanilla category is going to experience much innovation from this. There are a few limitations of this format. 1st, print quality vs manufactured quality. 2nd, ability to have focused iteration on a design to ensure maximum quality. 3rd, the literal centuries of combined experience of the engineers at Noctua can't be underestimated. They've been doing this for a Long time and have been the top dog for quite a while.
Why is the fan show down so compelling? I love it!
I find it more propelling than compelling.
for season 6 I think you should look at having 2 catagories.
the first where the fan is completely within the standard housing and
the open class where anything goes.
Just an idea.
I would like to see a return to extended vapor/smoke scenes for each fan in the episode. It kind of felt rushed to see all 4 with just 1 perspective thrown up on in the video. Plus individual fan scenes with both oblique and side perspectives allow the fan designer to identify how their fan is performing in your test and how they can possibly improve.
Just as an aside for the high-bypass and open fan designs for jet engines, there is a second limiting factor -- tip speed. The faster a fan rotates, the faster the blade tips move, and the larger the fan, the faster the blade tips move for the same rotation speed. It will likely never be a problem for computer fans, but the higher the mach number of a propellor or fan, the lower its efficiency becomes, and when the blade tips closely approach Mach 1, the noise produced goes up hugely. In the 1950s, there were two experimental XF-84H fighters built with a turboprop engine driving specially-designed supersonic propellors. These planes quickly acquired the nickname 'Thunderscreech', because the sound produced by the propellor when the engine was run up, in the plane of the propellor, was physically painful more than 100 yards away from the plane.
Thank you so much for not blasting the smoke test music in this video. My ears appreciate it. I love the fan showdowns. Keep them coming.
I love the videos and the 4K60 quality because every time I always click on the quality on my phone or PC and I always put it on the highest quality instead of stupid auto
I'd love to see the Turbo Pump intake, with the XL's "vector housing" (just the part after the fan)
Loved all the fans. Would be cool to see them do the smoke test 1 at a time and then all 4 at the same time... Builds up the anticipation :)
Nice episode Major, super interesting fans. Going back to the previous episode with the all-time best fans, I think that you could probably make that into a mini-series. Maybe you could run more tests such as radiator thermals, air cooler thermals, noise, and maybe a noise to performance metric. Then maybe another episode testing fdm vs resin. I know you have done comparison videos in the past, however, I feel like different designs could respond differently based on wing thickness and such, which could drastically alter the weight, and thus performance of the fan. Anyway, Keep up the great work, I'll never get tired of this series.
Oh and one more thing that I've been meaning to discuss with the community. There seems to be a missing metric which is never discussed in these videos, and that the cfm testing leaves absent. The turbulence of the air, the angle of the exhaust and it's rotation, as well as the speed, within the first three inches behind the fan, can be a significant variable that is overlooked in the current testing methodology. Also, the positioning of the exhaust can be a significant variable, especially with air cooling as the exhaust being concentrated closer to the vapor tubes can substantially change thermals. Because of this, I hope you consider going back to the previous testing methods of just checking thermals using a well built air cooler in the future, as I represents a much more real world performance metric, than just check cfm and static pressure. I understand the community likes the current testing method, but it does seem to fall short for me. I think that the information gathered from the wind tunnel is a nice piece of supplemental information, but it doesn't tell the whole story.
anyway, that's my two cents. I'd love to hear what the community has to say. And congrats on being a new parent!
As some other comments have mentioned, it seemed the XL might have been dragging on the housing. In the smoke test it seemed to be sagging quite a bit, possibly throwing everything out of alignment. Perhaps test again with a "kickstand" to line everything back up?
I love this series, my only complaint is that I would love to try some of these designs but the files are not available for download. It would be nice if there was a store where we could buy all the designs and the designers would get paid.
Wow! The Radish had a very very nice result in the smoke test! That looked like a very compact air flow.
Amazing Major Hardware!!!i like so much your videos!!
Missed this! Cool designs
An idea is on the XL instead of just having a drivetrain straight to the main fan you could also make another fan before that on the main part of the Noctua like normal and have that push air into the bigger fan which could help.
You have achieved that each time i see that you have uploaded a new video, i feel pure sincere childish joy. Thanks dude.
That tri one works best at approx 4000rpm. So it’s kind of impressive that it did that well at half speed.
I've been waiting for this since I first saw the propeller.
The XL would probably do much better without the ring on the fan, and with a stator. Probably still not super high performance, but it have the potential to be very quiet, if it doesn't rub, and with a silent bearing. One of the easiest ways to make a fan quiet, oversize it.
Rings on the fans is a bad idea, that's why it's (almost) not a thing. It hurts aerodynamics, pretty bad, and cause noise. I tried it when I made a replacement for my table fan, I knew it wouldn't have great performance, but I was really surprised by how bad it actually was. Very low performance, high noise and high power draw. If stretching of the blades is the concern, thicker blades is a much better solution.
Not having a stator on axial fans that push air into an open chamber is another classic, it's amazing how little back pressure it takes to reduce the flow a lot.
Besides those two issues, and the fact that the size makes it "not very practical" I really do like the design.
You put the Raddish Stator/Diffuser on the other side of a heatsink or radiator or a shroud to mount on a rad/inlet may help
I've seen numerous PC modders make videos "testing" a toroidal propeller as a PC fan. The whole point of a toroidal propeller is to prevent/reduce tip vortices - air or water wrapping around the tip of the blade from the high pressure side to the low pressure side, which reduces efficiency. This is completely unnecessary in a PC fan because PC fans are ducted. The ends of the blades attach to a shroud or duct, and the duct completely blocks these tip vortices.
Ducts are generally undesirable in boat or plane propellers because the extra surface area creates additional drag for the propeller, reducing its efficiency more than the efficiency gain for preventing tip vortices. Which is where the toroidal propeller has an advantage, because it can prevent tip vortices for a much smaller increase in drag (it has more propeller surface area, but not as much as a duct). Usually you only see ducted propellers in applications where objects entangling in the propeller are a risk. But if you're trying to blow air into/out of an enclosed space (like a PC), the edges of the fan blades are already going to be right up against a stationary object, so it is by default ducted. And there's no benefit to reducing tip vortices because there are none in the first place.
To be honest, the Turbo is probably the best out of all four.
The Radish has a little bit more airflow, but its all concentrated into the middle (Because its meant for thrust not cooling), where the Turbo is slightly less airflow, but spreads it out across more of your heatsink. Remember, Surface Area is as if not more important than airflow.
you could have left the cone bit on and done pull. It would have been louder than push but it looks like the blade bit fits in the housing
Man those resin prints look so good!
I am still curious about how the other brands fans work on the noctua hub. Especially the Arctic fan. So you get an even more equal comparison between them.
I love the ingenuity on the XL.
I strongly suspect, although I don't have a way to prove it, that the reason the Sharrow toroidal propellor comes up short as a case fan is the medium. Water is largely incompressible, so a propellor that moves larger quantities of water has greater thrust by direct reaction. Air, on the other hand, is readily compressible, so a fan that moves lots of air builds up a higher pressure behind the fan; given time for the air to expand again -- as in a high-bypass turbofan engine -- you recover the energy from that compression, but over the distances involved with a normal case fan setup, you're just getting a higher pressure behind the fan than you would with a more efficient fan design, not a greater airflow.
Have you thought about picking up a Phanteks T30? 30mm deep instead of the usual 25mm and a complete BEAST. I'm sure all these creative people would love to have another 5mm of depth in the fan frame to play with. ;)
This is what I was waiting for!
Now I want to see the XL with a more powerful motor, which is actually able to drive the fan.
Amazing content so far , but can you do tests for temperature change between intake and exhaust of the air. Maybe by creating a controlled air tunnel mimicking the inside of a PC case on the exhaust side.
That toroidal fan may be inspired by the one making all the news, but it's a completely different shape....
thanx for the great work !
The noise signature of turbopump was really nice!
The 80's music is back, thank you!
You should make a modified version of the radish which has a frame around that extra part to make it fit in front of the grill. It looks likely that it will be the overall winner if it's measured as intended.
I like it better than the prop. Work of art
I like both the narrow-ness of the Radish's exhaust flow, and the obtuse exhaust of the turbo.
If I were to use 120mm fans I'd throw the turbo on the bottom, and the radish in the center and top positions.
But 180mm fans is where it's at.
Seems very nice surface finish. I hope we can have the video of the printing process also.
Would have liked to see a spacer on the Radish during the test so that the second piece could be used.
I wonder how the Radish would do with the exhaust cone/stater add on on blowing through the radiator. Any possibility of adding a spacer for it in between it and that radiator that would allow for clearance of the cone/stater? I think that would be cool to see!
A dual fan episode would be fun where you mix the best high volume and high compression designs. I think the XL with the Radish mounted on the end easily beat the best single fan.
Also do you think the radiator should be in the middle or at the end?
Oh, nice concept!
Edit: sandwich the rad in the middle.
@majorhardware I don't know if you can take down comments but it looks like someone is impersonating you trying to scam your viewers. They responded to my comment and a couple others.
kudos for showing all the fans pushing through the radiator. The Noctua is still impressive and I forgot that the Cheater with a radiator is not the Cheater king.
Oh noooo, with that 3rd design i can imagine everyones gonna try making bigger fans now lmao but not taller i hope xD
dang its crazy and pretty cool that the a12x25 still has not been beaten. It may not ever be, but if that day ever comes, it'll be cool to see what the cooperate world does about it. Good show dude
False I think it was in season 2 or 3 someone got a brand deal for beating the goat
The really insane thing is that the A12 is largely inspired by the Nidec Gentle Typhoon, that was released like 15 years ago. The A12 is better, but it's also using better materials and manufacturing methods. The GT's only cost ~$13, and that's accounting for inflation.
@@alexanderdesfosses I remember the brand deal because of the design, but I can't remember if it actually beat the a12. Also the a12 is still at the top of the list so I just assumed it wasn't beaten
@@MlnscBoo Well, older seasons used different testing setups that prioritized pure airflow, so some designs did better than the A12 in that environment. Now that we're measuring the air that can be pushed through a radiator, the A12 is showing off what it was optimized to do, and hasn't been beaten. And yeah, iirc, the fan that got the brand deal was not necessarily because it beat the A12 in performance metrics, it was because it was *pretty* good and also had a very cool serrated look.
Those 3d prints look really good
The xl is my personal favorite, it looks so cool
play WoT till tier 7 vehicles if you choose the German line , from 8 you have to both win AND survive the match , otherwise you're going down on profit . i loved the VK100 but losing money ( as if losing many matches wasn't enough ) was too much . Medium and Heavies , i recommend ONLY the American line : very good front armor , especially turret ; very good gun depression ... T29 is also a great Heavy , M6 is also worth keeping too . if you choose to buy premium tans : T26E4
It's so charming how all the DIY engineers make their fans... Just like in cargo-cult make "look alike thing" and it will work somehow. If I need to design fan, I'll analyze flow I need, power I have, and just calculate blade configuration for this exact case. Modern fans can produce up to 97% of efficiency, so you can't go against physics and invent over 100% ECE configuration. However, you can create a little less noisy fan with some design efforts.
Best episode so far
Omg the prints so much better
My outside-the-box idea is, if anyone wants to try and create it for me, is what I instantly came up with before the XL was shown (after he hyped it):
It's like a high bypass, in essence.
Inside the hub is a high speed turbine style (aka jet), like on _Turbo Pump;_ *however,* instead of the velocity stack, it extends out BEYOND the frame of the fan and transitions the pitch into a 200mm Ultra-High Bypass sickle style blade. _(it will probably need a support rings @150mm, and blade tip; aka 200mm)_
Then, to to it all of, is a big shroud that has a velocity lip and funnels the 200mm back down to 120mm... ☺️😏
Therefore combining the airflow from the inner high speed turbine PLUS the high volume from the UHBP portion.
My hope, in the end, is that _if nothing else_ the high volume airflow will act as a "lubricant" and cause air to channel inwards at an even higher speed, thanks to the inner turbine blade! 🥴
_[I don't really expect anyone to try this, but 🤞🤞❤]_
Those top few behind the x25 on the leaderboard might deserve a rerun in resin at some point. Given the difference we saw between print qualities they might have a chance of coming out on top.
The inlet runner needs to be longer so it'll accelerate the air before hitting the blades. I've been able to improve CFM by installing a 4" duct adaptor on the inlet end of a 120mm fan.
love the blue print colour on this ep :)
Radish airflow is awesome-looking
I like Rodrigo's out of the hub thinking, also in editing pls show the image/picture of the fan in the corner during the soundtest
It is impressive how well he is performing the Noctua!
Love that blue filament colour
The best part of WoT is that if you hang around on the forums for long enough you'll get access to top secret information.
As a Stephen, Stephen = Steven. I still don't understand how this isn't known by everyone.
The radish is an embodiment of "slow is smooth and smooth is fast". It works more efficiently and thusly harder with less effort.
I think the radish scores higher because the fixed fins reduce turbulence allowing the air it rejects to be sent further. It doesn't mean it's any more efficient
we could use a real scan of the marine prop so we can trial it in RC Boats this tech is big news great job thanks
I will give you one nice idea about how air differs from water - it is much lighter, so you can "throw" it. In other words, you can involve air mass with showel like, scoopy fan. But always remember that air will be pushed tangent to blade plane
Good example - termalright old fans.
Big scoopy blades good with flow, but less with pressure
Nice, have you ever thought about a cyclone dust filter for PC fans? I got one office where the ordinary filters need doing once a week!
8:48 so the Turbo Pump is the most powerful one which is actually quieter than the reference? And that by a long shot - nearly half the noise level.
...Rodrigo was really thinking outside the box by creating a box of his own...:D
you could have added a spacer for the jet engine design for the cone to be used. Just an idea if you retest it.
Well, now I really wanna see the raw airflow from the XL!
They look so cool. No matter the performance I want all of them in my bulid ! :D
I had the same idea as the XL, but i didn't submit it because i knew the motor couldn't drive it. A 120mm version would probably work much better
Yes, I agree with you. You can see the RPM is very low, meaning the power the electric motor is putting out is very low. It's running on half the power that other fans are. Hope you give it a try, but reducing the angle of attack of the blades, and the blade area will reduce the "bite" of air, allowing the motor to spin faster and the motor to use more power. More Power, more air flow. It's just like having the wrong propeller on a boat motor. The motor needs to spin at a certain RPM to maximize power.
Great vid.
The XL was spoilt by the ridiculous ring attach to the ends of the blades which does nothing but add rotational mass that consumes power.
Enlarge the cheater blades or 1 of the other past ringless fans that ranked well, to fit snugly into the XL enclosure instead of the hideous ring fan.
Regards, Jas.
VK4FJGS
Rocky Qld.
I'd love to see the XL hooked up to Dremmel could possibly be a great lef blower 🤣
Can you try printing with ultralight pla? See the difference for the XL fan perhaps, or test the same fan with different pla in general? Maybe the motor isnt up to the task in some fans :)
First time I saw that propellor come up I was hoping you'd turn it into a PC fan XD
Triskelion is a Runescape reference. Triskelion keys are a three section key for a chest. The three pieces make a circle key.
I love that you don't want to buy the boat propeller, but you're surrounded by a hundred thousand dollars worth of 3D printers
Some get gifted to him as he does product reviews
do u have a high temp resin? u could probably print a small turbo for a lawn mower if so, since the exhaust temps are generally low (for exhaust at least) would be cool to see.
The Sharrow prop concept does not have any benifet in a ducted fan scenario. This has already been tested in marine conditions with bow thruster tests (by Sharrow themselves). I don't think there's much more to be gained, without using a non pc fan format for mounting. The prop has to be separated from the motor by a long shaft to take advantage of the fluid dynamics involved.
Apparently Noctua knows what they're doing. I always use Noctua or Delta. Delta in servers and Noctua's in desktops. The triskelion was really interesting.
Errr... I could not watch this video... It was just... I don't know... I LOVE !!!!! ALL you other videoes but this one... Broke my brain!!!!
2:02 LOL 😂 Yeah, that's a pretty high by-pass.
I really like the XL it does what it looks like it should do