Airflow vs static pressure fans. Are they a scam?

[doyll]

Hi, @doyll!
A good fan has to have a good "jet", but not a lot of "wake" which causes turbulence. Wake is very detrimental to a good fan. Too much wake and a fan starts working in static pressure, making just noise and not a lot of air flow. We should look at its pq curve where it is just below its static pressure build up point.

I tried to simplify it as much as possible.

Basically "airflow" is all about displacement .. air moving from higher pressure are (at fan) to lower pressure area (farther away from fan.
We can't move more air into our cases than is moving out of it.

Our fans have such low static pressure ratings even at full speeds of 1300-1500rpm. They make less static pressure / pressure differential than the difference in pressure on our chests vs on our feet standing at sea level) at full / maximum speed. As most of us run our fans at 700-1000rpm (1/2-2/3rd of their full speed) lowers their pressure rating to about half of spec .. so about the smae as pressure on our butt vs on our feet. That is extremely little ability to overcome any kind of resistance in our cases (grill, filter, cables, etc).

Therefore we need fans with as high a pressure rating as possible at the speed we will be running them.

Like I said, this is over-simplification. While I understand your "jet" and "wake" concept, they have little to no effect in our cases with HDD cage, GPU PCB and cooler, cables, RAM, etc. all disrupting "Jet" & "wake" effects.

I'm trying to get some equipment and samples together to compare how different impeller designs effect case airflow. My guess is impellers like NF-A12x25 and TL-C14 (Thermalright) will give us better airflow through our cases than impellers like NF-A15 and TY-147. Hopefully there will be enough difference it will show in test results.

 

[mtcn77]

TY-147

Lol, Thermalright 147 can eat those 12cm fans for lunch. Take a look at "hardware.info".

While I understand your "jet" and "wake" concept, they have little to no effect in our cases with HDD cage, GPU PCB and cooler, cables, RAM, etc. all disrupting "Jet" & "wake" effects.

How good of you to note, however jet and wake flows form the bulk of fan volute, so if it can't eject useful airflow, wake currents arise.

My personal take is that pressure optimised fans create a smaller volute, albeit a more resistant volute flow seperations due to wake formations.

 

[The red spirit]

@The red spirit good stuff, but he took cfm at face value, not at optimal fan efficiency pq value.

I think that bigger downfall of that post is that fan spec data collection methods are undisclosed and they are likely aren't accurate. I dunno, I start to feel that generic 7 blader is one of the best fan shapes and everything else is trying to deviate from it for deviation's sake.

I'm trying to get some equipment and samples together to compare how different impeller designs effect case airflow. My guess is impellers like NF-A12x25 and TL-C14 (Thermalright) will give us better airflow through our cases than impellers like NF-A15 and TY-147. Hopefully there will be enough difference it will show in test results.

If you decide to do any testing, please test generic 7 blade fan.

 

[mtcn77]

I think that bigger downfall of that post is that fan spec data collection methods are undisclosed and they are likely aren't accurate. I dunno, I start to feel that generic 7 blader is one of the best fan shapes and everything else is trying to deviate from it for deviation's sake.

Total blade surface area can benefit from a lower blade count, considering fewer impeller blades have directly correlated less total trail circumference. The whole of the blade does not generate useful work. A bigger blade can increase the suction zone behind the leading edge, from what I gather air only moves to fill the atmospheric pressure behind the fan blade.

Just look at the plateau under the static pressure peak in pq curves. Its the most effective fan impedance curve point. It is both reliable and not very outlandish in cfm rating, basically ~%50.

 

[cvaldes]

smaller dia fans are generally poor performing due to them having a bigger ratio of the hub sacrificing wind swept volume.

it's interesting to see noctua have 2 very different fans with one optimised for airflow and the other for static pressure at the same rpm but not the same noise profile? seems to me that AF is noiser than SP??

View attachment 207544View attachment 207546
View attachment 207547
View attachment 207548

the SB series is louder, less airflow and more pressure than the SA series, interesting!

Noctua has three primary fan designs not two.

NF-A__ - all-purpose, 9 blades
NF-F__ - static pressure optimized, 7 blades (these are the NF-P__ products in the redux line)
NF-S__ - airflow optimized, 7 blades

The S12A and S12B are have the same blade design however Noctua selected different parts (plastics, fan components) and subtracted a few details that makes the less-expensive redux-1200 product line slightly noisier. That's why they are so similar in performance.

Noctua themselves explain their fan differences in several articles:

Which fan is right for me?

Designed in Austria, Noctua's premium cooling components are renowned for their superb quietness, exceptional performance and thoroughgoing quality.

noctua.at

NF-A12x25: performance comparison to NF-F12 and NF-S12A

Designed in Austria, Noctua's premium cooling components are renowned for their superb quietness, exceptional performance and thoroughgoing quality.

noctua.at

Noctua

In addition to our standard line-up of premium quality quiet cooling fans that feature all our latest technologies, come with a complete set of accessories and are easily recognisable by our signature brown colour scheme, we carry two extra product lines: redux & industrialPPC.

noctua.at

Following Noctua's suggestions I use NF-F12 PWM fans on my radiators and NF-S12 PWM fans for chassis intake/exhaust in three PC cases. With my current fan curves, the fans all top out around 1000 rpm, well short of the maximum rotational speed (and thus noise). One build uses the grey NF-F12 fans on the radiators for aesthetic reasons. I don't actually own an NF-A12x25 all-purpose fans since I am using task-specific models.

In my NZXT H1 case I replaced the stock NZXT Aer 140mm fan with the NF-A14 industrialPPC-2000 PWM. Paired with a Low Noise Adapter cable and again with a fan curve that doesn't extend to the fan's maximum speed, this is considerably quieter than the stock fan.

In your table, you fail to recognize that the Low Speed variant of the NF-A12x25 provides barely half the airflow as the S12 models. So yes, the NF-A12x25 LS-PWM is quieter at full speed than the S12 it's moving considerably less air. A more realistic real world scenario would have the S12 fans running between 50-75% max speed which does knock down the noise substantially.

 

[mtcn77]

Noctua has three primary fan designs not two.

NF-A__ - all-purpose, 9 blades
NF-F__ - static pressure optimized, 7 blades (these are the NF-P__ products in the redux line)
NF-S__ - airflow optimized, 7 blades

The S12A and S12B are have the same blade design however Noctua selected different parts (plastics, fan components) and subtracted a few details that makes the less-expensive redux-1200 product line slightly noisier. That's why they are so similar in performance.

Noctua themselves explain their fan differences in several articles:

Which fan is right for me?

Designed in Austria, Noctua's premium cooling components are renowned for their superb quietness, exceptional performance and thoroughgoing quality.

noctua.at

NF-A12x25: performance comparison to NF-F12 and NF-S12A

Designed in Austria, Noctua's premium cooling components are renowned for their superb quietness, exceptional performance and thoroughgoing quality.

noctua.at

Noctua

In addition to our standard line-up of premium quality quiet cooling fans that feature all our latest technologies, come with a complete set of accessories and are easily recognisable by our signature brown colour scheme, we carry two extra product lines: redux & industrialPPC.

noctua.at

Following Noctua's suggestions I use NF-F12 PWM fans on my radiators and NF-S12 PWM fans for chassis intake/exhaust in three PC cases. With my current fan curves, the fans all top out around 1000 rpm, well short of the maximum rotational speed (and thus noise). One build uses the grey NF-F12 fans on the radiators for aesthetic reasons. I don't actually own an NF-A12x25 all-purpose fans since I am using task-specific models.

In my NZXT H1 case I replaced the stock NZXT Aer 140mm fan with the NF-A14 industrialPPC-2000 PWM. Paired with a Low Noise Adapter cable and again with a fan curve that doesn't extend to the fan's maximum speed, this is considerably quieter than the stock fan.

In your table, you fail to recognize that the Low Speed variant of the NF-A12x25 provides barely half the airflow as the S12 models. So yes, the NF-A12x25 LS-PWM is quieter at full speed than the S12 it's moving considerably less air. A more realistic real world scenario would have the S12 fans running between 50-75% max speed which does knock down the noise substantially.

Extreme fan designs do not pick up in consumer space for a reason. The demand is driven by all around laminar performance. It is hard to design a fan with no stall and as much performance as possible, yet the use case is very much limited to that.
We don't need too much "terminal" air velocity in order to avoid trail separation. What I have been thinking about is if Noctua launched these fans in opposite flow direction, so we could test out their different designs in series like vane axial fans working together, but different push-pull units.

 

[cvaldes]

You can wish all you want but it is highly unlikely Noctua will be releasing opposite flow designs in the near future. The market for that item is exceptionally small; most PCs sold are pre-built.

DIY builders are already a small subgroup of PC users and many are happy just buying the cheapest aRGB model. The performance focused enthusiasts are rare.

 

[80-watt Hamster]

You can wish all you want but it is highly unlikely Noctua will be releasing opposite flow designs in the near future. The market for that item is exceptionally small; most PCs sold are pre-built.

DIY builders are already a small subgroup of PC users and many are happy just buying the cheapest aRGB model. The performance focused enthusiasts are rare.

Noctua may not, but Antec does.

 

[mtcn77]

Noctua may not, but Antec does.

The brands don't even need to be identically matched, it is just to match them in alternative push-pull setups to see which works.
PS: this could work. Antec, you, pc-friendly brand!

 

[cvaldes]

Well then enjoy your Antec fan analysis.

I was just addressing your comment about wishing that Noctua would ship this. They aren't.

That said, I presume that Noctua has hundreds of prototypes of unreleased designs in their labs. Only a handful make it as shipping product. All of the good manufacturers do this: Apple, Samsung, HP, Lenovo, et al.

In fact, creating prototype opposite flow designs likely helps them design future versions of their products that ship today. Many of Noctua's current fans are iterations of older designs.

 

[mtcn77]

DIY builders are already a small subgroup of PC users and many are happy just buying the cheapest aRGB model. The performance focused enthusiasts are rare.

You know, in my religious text, the literate are the few, yet the stronger of the bunch. Don't fret.

 

[cvaldes]

Yes, DIY PC building is a religion.

That said, some are aggressive missionaries, others are Zen Buddhists.

 

[mtcn77]

Yes, DIY PC building is a religion.

No, becoming a believer goes through being literate. Don't worry about which products "we" can make cool and trendy. Idiots follow wherever we go.

 

[cvaldes]

No, for thousands of years believers of many religions all over the world were illiterate. In fact, it was helpful to keep your faithful largely ignorant. Thoughtful analysis and critical analysis of standing dogma results in disagreement.

But do enjoy those Antec fans.

 

[mtcn77]

No, for thousands of years believers of many religions all over the world were illiterate. In fact, it was helpful to keep your faithful largely ignorant. Thoughtful analysis and critical analysis of standing dogma results in disagreement.

But do enjoy those Antec fans.

Not if you go look at what is written which isn't a matter of opinion. It is okay to hold the opinion that it is obfuscated by ambiguity, but it is very specific about the requirement of literacy and true belief being a virtue of that.
Like I said, until someone does it first, it cannot be trendy to do so, like believing it can be done.

 

[cvaldes]

That's fine.

I just care that my builds are cool and quiet. I have accomplished this with my cooling implementation.

I can run a synthetic benchmark like Cinebench R23 or a real-world task like a Handbrake encode and my Ryzen 5900X remains cool and quiet. Same with my TUF Gaming GeForce 3080 under the Unigine Heaven benchmark or during actual game play.

This is a solved problem and didn't require rocket science. I don't need a Ph.D. in aeronautical engineering or hydrodynamics. I do hope fan engineers have those degrees though.

If people want to make this more complicated than necessary, they are free to do so. I have better things to do with my time.

 

[mtcn77]

I just want to know rotor terminal velocity in correlation to axial velocity of air. Since air behaves with differences in velocities differently, 2 fans could accelerate it better than any one single design and not exceed the separation terminal velocity.

Okay, so... pumps are not any different than air cooling fans, what is next...

 

[The red spirit]

If people want to make this more complicated than necessary, they are free to do so. I have better things to do with my time.

Sorry mate, but in this thread there's a desire to find out why fans are the way they are and why that matters (or doesn't).

 

[Shrek]

Laptops use centrifugal fans, but are there any such fans for desktop use?

 

[The red spirit]

Laptops use centrifugal fans, but are there any such fans for desktop use?

Blower GPUs? This thing:

This Dell:

Cooler Master Jet:

 

[cvaldes]

The centrifugal fan design is inferior in airflow which is why it's primarily used where space is at a premium. If I recall correctly, this style of fan is also noisier.

Some GPUs have blower-style fans because it is expected that the card will be deployed in a system that lives in a climate controlled server room where noise isn't an issue. Usually those are cards that are designed to be installed in 2 PCIe slots so having a louder cooling solution is preferable to eating up more real estate.

 

[freeagent]

I had that Jet 7+ on my old axp-m 2500. It made a whistle sound just off idle so about 17-1800rpm. It was pretty decent. Lots of copper on a bare die was helpful with that insane clamping force lol.. The advantage was supposed to be no dead spot.

 

[mtcn77]

Some GPUs have blower-style fans because it is expected that the card will be deployed in a system that lives in a climate controlled server room where noise isn't an issue.

Actually, this type of forward inclined radial fans are self cleaning compared to backward inclined fans which surprisingly have a higher efficiency, if I recall correct. Also, they are quiet compared to forward inclined radials. Their only handicap is they are sensitive to dust wearing their impeller.
PS: this thread could not have taken off as the only nut job being myself. As shown here, there are other people with fan daydreams.

 

[The red spirit]

I had that Jet 7+ on my old axp-m 2500. It made a whistle sound just off idle so about 17-1800rpm. It was pretty decent. Lots of copper on a bare die was helpful with that insane clamping force lol.. The advantage was supposed to be no dead spot.

But the main question is, did it sound like jet?

 

[freeagent]

But the main question is, did it sound like jet?

It did kind of lol! That whistle was annoying AF. I still have it in a box somewhere lol..