Airflow vs static pressure fans. Are they a scam?

[80251]

This one blade 'fan' is balanced

I'd like to see that thing fly -- from a safe distance.

You'll notice that by the end of the war all the high-performance, allied, single, piston-engined fighter aircraft were using gigantic, paddle-bladed, 4-prop, steel propellers.

Aren't those heavy, asymmetric, direct driven, metal radiator fans a thing of the past? It seems like most automotive radiators now use lightweight electric fans.

 

[agent_x007]

Because those high CFM fans spin at crazy rpms and are loud, if not outright burning every motherboards' fan connectors. You know, they have maximum power ratings too and they surely aren't made to handle 20+ watt fans.

Sorry for quality, it's video from 6 years ago (fans stil work no problem in this configuration) :

Short version :
Get a PWM high speed fan, pull 12V and GND from it's connector and stick it into Molex 12V and one of it's GND (leaving PWM and FanTach to MB).
Profit
PS. I exchanged push fan for TFC1212DE in 2017. In 2015 (on this video), it was still FFC1212DE.

 

[The red spirit]

Short version :
Get a PWM high speed fan, pull 12V and GND from it's connector and stick it into Molex 12V and one of it's GND (leaving PWM and FanTach to MB).
Profit
PS. I exchanged push fan for TFC1212DE in 2017. In 2015 (on this video), it was still FFC1212DE.

I don't want that. Why would I ever want such a noisy PC? My CPU fan works within this rpm range 300-1200. Pretty much never it reaches maximum RPMs. A bit different subject is RX 580, because I can't change its cooling and PWM logic, but I can adjust power limit (and some other things) in vBIOS. Anyway, that's off-topic, all I want to know here is why fans perform the way they do.

 

[mtcn77]

I'm sure the issue why some fans can't push the radiator air is settled. It is okay to have preferences, but those preferences doesn't always reflect their sound profile consequences. Otherwise, we would see Noiseblocker eLoops dethroned by other fans - some with stator rings more centrally placed in the rotor frame, or some even without it. Not going to happen in truth. Sound vs. pressure balance takes no compromise.

 

[freeagent]

Why would I ever want such a noisy PC?

Because high performance and silence don't go hand in hand.

 

[The red spirit]

Because high performance and silence don't go hand in hand.

Do you honestly think that someone like me, who own i5 10400F and detuned RX 580 really cares about performance much? Anyway, can this shit end? This is off-topic and if you wanna say something, then say something about fan airflow or pressure and why fans perform the way they do. This is interesting, not some server equipment.

I'm sure the issue why some fans can't push the radiator air is settled.

Sort of. It still bothers me that the only "push through radiator" metric we have is static pressure. Fan spins, it's not static and radiator isn't completely blocking airflow. It still doesn't make sense to me to use static pressure, instead of pressure at different rpms, as well as airflow at different rpms.

It is okay to have preferences, but those preferences doesn't always reflect their sound profile consequences. Otherwise, we would see Noiseblocker eLoops dethroned by other fans - some with stator rings more centrally placed in the rotor frame, or some even without it. Not going to happen in truth. Sound vs. pressure balance takes no compromise.

I actually wonder how would eLoops perform at 3000 rpm, just for data.

 

[mtcn77]

Fan spins, it's not static and radiator isn't completely blocking airflow. It still doesn't make sense to me to use static pressure, instead of pressure at different rpms.

When you look at a p/q curve, it is actually an impedance/q curve. You cannot set the radiator fin gap by alternating the fan rpm. You have grown accustomed to thinking you can fix it. It is particular to the fan characteristic.

I actually wonder how would eLoops perform at 3000 rpm, just for data.

Not that well. They can enter the charts at ~1500rpm, they aren't particularly good when noise is present, stator rings have their own frontal section which block free flow of air at high velocity. They just meet turbulent area with efficient operation area better. They aren't case fans per se.

 

[The red spirit]

When you look at a p/q curve, it is actually an impedance/q curve. You cannot set the radiator fin gap by alternating the fan rpm. You have grown accustomed to thinking you can fix it. It is particular to the fan characteristic.

Still, so many sources call that static pressure. And those charts are made at fixed rpms.

Not that well. They can enter the charts at ~1500rpm, they aren't particularly good when noise is present, stator rings have their own frontal section which block free flow of air at high velocity. They just meet turbulent area with efficient operation area better. They aren't case fans per se.

But at low rpms, they are kinda good. Say what you want, but they look really cool.

 

[mtcn77]

But at low rpms, they are kinda good. Say what you want, but they look really cool.

That is the point. Every fan until its impeller's stall speed is good. They also don't work that well in a pull configuration, that does not bring down their value one bit. It is just designed the way it is.

Still, so many sources call that static pressure. And those charts are made at fixed rpms.

You should look inside the area of the curve. The rpm ranges make impedance variable efficiency peaks that form an incline inside the p/q curve. Rpm just changes the incline, the stall point to radiator restriction stays the same. Changing the rpm won't make it a pressure fan per se.

 

[The red spirit]

That is the point. Every fan until its impeller's stall speed is good. They also don't work that well in a pull configuration, that does not bring down their value one bit. It is just designed the way it is.

That's because they are static pressure fans. As I understand good static pressure fans are always good for pushing air through stuff, but may not be so great at pulling.

You should look inside the area of the curve. The rpm ranges make impedance variable efficiency peaks that form an incline inside the p/q curve. Rpm just changes the incline, the stall point to radiator restriction stays the same. Changing the rpm won't make it a pressure fan per se.

Really?

It show that upping rpms, does indeed improve pressure and airflow. Stall point just shows that at certain impedance fan will be working sub-optimally, meaning that it will be in poorer pressure-airflow ratio at same rpm.

 

[mtcn77]

Stall point just shows that at certain impedance

Yes, at certain restriction a.k.a front section "aerodynamic drag," the fan will start to stall the same as it does in a pull configuration.

These fans cannot break aerodynamics. If they accelerate air more than a certain limit, turbulence begins and voids happen in the impeller tail and front flow. If there weren't any pressure change, there wouldn't be any sound. All development is to fall within these limits, not break them.
PS: the fan you demonstrate is pretty good by the way. It is self-correcting which makes me guess it is backward curved.

 

[The red spirit]

Yes, at certain restriction a.k.a front section "aerodynamic drag," the fan will start to stall the same as it does in a pull configuration.

I meant it more alongside of air filter in front of it, closed doors or it blowing on radiator or cooler. I didn't mean frontal area. Frontal are mostly doesn't mater for consumers fans for simple reason, because it becomes a restriction at higher airflow situations. Most consumer fans don't spin that fast for frontal area to matter, however, I referenced 9800 rpm server fan and it had only three blades and certainly didn't have spinning disc around it like eLoops.

When I was reading about other fans today, I found out that other fans and fan alike things (like planes) have less blades, when they are engineered to spin very fast and move a much air, when there's no obstruction. Imagine something like aircraft 3 blade propeller or ceiling fan. More blades are added, when fan is designed to have more pressure. More blades is more pressure, but less maximum airflow. You can also make flat blades and turns out that they are better at moving more air, but at cost of noise. So what I found out, is that computer fans are mostly very similar to each other, is due to specific requirements for it. Generic 7 blader is the way it is is to provide a adequate amount of pressure and provide decent air flow at, relatively to other things, low revs and low noise level. 7 blades gives it pressure, non flat blades reduce noise, and while frontal are of it isn't small, it still provides okay air flow. It's very hard to actually increase airflow of fan due to noise restrictions of computer users. There's not much of what you can do. You can make fan spin faster or grow in diameter. Most fancy fan designs add more blades, make blade shape different and that's due to them trying to figure out how to reduce noise or to increase pressure. There clearly aren't many pure airflow oriented designs and for good reason, because computer is restrictive and airflow oriented designs just wouldn't work, due to their lack of static pressure or due to their higher noise. All things considered, eLoops in theory should be great, if you want quiet and pressure oriented fans. Meanwhile something like Noctua S12 series should be good as high airflow fans, maybe at slightly lower noise levels, but only if they are used with minimal restrictions otherwise it will perform worse than even the generic 7 blader.

PS: the fan you demonstrate is pretty good by the way. It is self-correcting which makes me guess it is backward curved.

Sorry, I have no idea what fan is referenced in this illustration. I found it on rage3D forum, when some user commented that computer fan manufacturers don't provide enough data. That makes me think, that it may not even be a computer fan or it is some industrial fan. Considering that it tops out at 2100 rpm, it may be some low end server fan and it's most likely generic 7 blader, because most server fans are generic 7 bladers. Only small fans or super fast fans aren't 7 bladers.

 

[mtcn77]

I meant it more alongside of air filter in front of it, closed doors or it blowing on radiator or cooler. I didn't mean frontal area. Frontal are mostly doesn't mater for consumers fans for simple reason, because it becomes a restriction at higher airflow situations. Most consumer fans don't spin that fast for frontal area to matter, however, I referenced 9800 rpm server fan and it had only three blades and certainly didn't have spinning disc around it like eLoops.

When I was reading about other fans today, I found out that other fans and fan alike things (like planes) have less blades, when they are engineered to spin very fast and move a much air, when there's no obstruction. Imagine something like aircraft 3 blade propeller or ceiling fan. More blades are added, when fan is designed to have more pressure. More blades is more pressure, but less maximum airflow. You can also make flat blades and turns out that they are better at moving more air, but at cost of noise. So what I found out, is that computer fans are mostly very similar to each other, is due to specific requirements for it. Generic 7 blader is the way it is is to provide a adequate amount of pressure and provide decent air flow at, relatively to other things, low revs and low noise level. 7 blades gives it pressure, non flat blades reduce noise, and while frontal are of it isn't small, it still provides okay air flow. It's very hard to actually increase airflow of fan due to noise restrictions of computer users. There's not much of what you can do. You can make fan spin faster or grow in diameter. Most fancy fan designs add more blades, make blade shape different and that's due to them trying to figure out how to reduce noise or to increase pressure. There clearly aren't many pure airflow oriented designs and for good reason, because computer is restrictive and airflow oriented designs just wouldn't work, due to their lack of static pressure or due to their higher noise. All things considered, eLoops in theory should be great, if you want quiet and pressure oriented fans. Meanwhile something like Noctua S12 series should be good as high airflow fans, maybe at slightly lower noise levels, but only if they are used with minimal restrictions otherwise it will perform worse than even the generic 7 blader.


Sorry, I have no idea what fan is referenced in this illustration. I found it on rage3D forum, when some user commented that computer fan manufacturers don't provide enough data. That makes me think, that it may not even be a computer fan or it is some industrial fan. Considering that it tops out at 2100 rpm, it may be some low end server fan and it's most likely generic 7 blader, because most server fans are generic 7 bladers. Only small fans or super fast fans aren't 7 bladers.

Frontal area "drag coefficient" means all that restriction in front of the fan - filters, grills - and rear area drag means general air impedance - shrouds, radiators - stalling the fan, so eLoops don't work in pull configuration for the same reason any other fan doesn't. It is just air impedance costing you noise.

You can make the same fpm from a fewer blade count, but the trail velocity would have to compensate which would incur more turbulence at the trail edge.

I forgot that fans operate very linearly when in turbulent flow, so noise and performance, such as your example fan, don't mix all that well which brings us to the same conclusion about consumer fans being temperamental about drag coefficient.

 

[Shrek]

I'd like to see that thing fly -- from a safe distance.

Aren't those heavy, asymmetric, direct driven, metal radiator fans a thing of the past? It seems like most automotive radiators now use lightweight electric fans.

I just found it amazing that people took it to such an extreme; but asymmetric computer fans are very much things of the present.

I think the Russian's use counter rotating propellers, trading complexity for efficiency.

Howard Hughes crashed in an XF 11 with counter rotating props; I think a pitch control failure caused the props to be blowing in opposite directions.

 

[mtcn77]

I think the Russian's use counter rotating propellers, trading complexity for efficiency.

Not a good example. Like vane axials, they cannot compound each other when turning in the same direction, especially for pressure building.

 

[The red spirit]

I think the Russian's use counter rotating propellers, trading complexity for efficiency.

That's what Scythe Fuma 2 does too.

 

[Shrek]

Fascinating, but I suspect the price of counter-rotating fans is extra noise; but I could well be wrong.

 

[Ominence]

I use two Silencio FP fans at 1080 rpms

your Silencios seem a good pressure optimsed fit for your sensory deprivation cube experience. i found this link kinda interesting.

Table chart sound pressure levels SPL level test normal voice sound levels pressure sound intensity ratio decibel comparison chart conversion of sound pressure to sound intensity noise sound units decibel level comparison of common sounds calculation compression rarefaction loudness decibel dB scale ratio factor unit examples - sengpielaudio Sengpiel Berlin

i have old noiseblockers as well so the eloops discussion is also somewhat interesting. i tried an interesting experiment some 4 years ago- Noiseblocker black silent pro 140mm 1700rpm. But my 18 fans sound like this on idle, 7v i think, 1 inch distance. this is a push pull arrangement sandwiching a watercool 420 pro stainless rad.

 

[freeagent]

Do you honestly think that someone like me, who own i5 10400F and detuned RX 580 really cares about performance much? Anyway, can this shit end

I don’t know what equipment you have. I did see those sweet Delta’s though. Enjoy your adventure

 

[80251]

I don't want that. Why would I ever want such a noisy PC? My CPU fan works within this rpm range 300-1200. Pretty much never it reaches maximum RPMs. A bit different subject is RX 580, because I can't change its cooling and PWM logic, but I can adjust power limit (and some other things) in vBIOS. Anyway, that's off-topic, all I want to know here is why fans perform the way they do.

Thicker fans can spin slower and still produce more CFM and static pressure than a thin fan -- at the same RPM. Whether or not it would also produce less noise is anyone's guess.

 

[Shrek]

I think it is fair to say that slower is quieter, although shape plays a role.

 

[Ominence]

hopefully we are all talking terminologies more correctly than the other way round.
fan diameter and fan thickness should not be confused.
in a fan advertised as being 120x120x25 in the specs, the fan dia is 120mm (the swept area) and fan thickness is 25mm which then leads to the swept volume calculation of the fan.

 

[mtcn77]

hopefully we are all talking terminologies more correctly than the other way round.
fan diameter and fan thickness should not be confused.
in a fan advertised as being 120x120x25 in the specs, the fan dia is 120mm (the swept area) and fan thickness is 25mm which then leads to the swept volume calculation of the fan.

Obviously, the more blade area there is, the more spherical the blades are in shape. Other than that, shaping them in constant overlying flow acceleration is the practical limit, imo. I think altitude matters for this scenario, but cannot tell in which direction. We would have split blades if it came to that.

 

[The red spirit]

Fascinating, but I suspect the price of counter-rotating fans is extra noise; but I could well be wrong.

They aren't any more noisier than any other fans. Come on, it's Scythe. They are specializing in quiet and powerful cooling solutions far more than other brands. Anyway, if yo uare interested in Fuma 2, you have to watch these two videos, they explain everything:

TL;DR is that it performs great and is very quiet, it's also great value for money. It performs like highest end dual tower coolers (minus 140mm models) and is at top or near top of quietness charts. Mostly due to the fact that Scythe just puts rather slow fans on it.

 

[Shrek]

Well, I don't quite consider this straight counter rotating as the air is straightened out in-between by fins.