Recently i took a task of replacing my stock case/AIO fans for something better and more quiet. As probably great majority of people i started browsing websites and sort fans by static pressure and airflow because that's what tells how good a fan is, right? You couldn't be more wrong. Thanks to some smart people on the internet i realised how big and massive mistake is it to even look at the CFM and mmH20 airflow and pressure specs listed on most of websites presenting PC fans. To spread awareness i made this post, in hopes nobody will ever make same mistake as me.
 
If you want to read, go on, if you want to listen, i made a video that explains everything https://www.youtube.com/watch?v=Ccu7mcJPbb0

First of all, the Airflow and Static Pressure specs listed at fan sheets over at shops selling fans and even manufacturers who publish product pages are Maximum Airflow and Maximum Static Pressure (some manufacturers will omit the "maximum"). So what's wrong with that? The thing is that:

- Maximum Airflow is measured when there is 0 resistance (fan does not need to generate pressure), meaning this is a situation when you put a fan vertically on top of you desk and let it blow - situation like that, is completely unrealistic and is NEVER happening in real life usage of PC fans. Even a simple case grill, a filter will create resistance on the path of air and thus the listed "Maximum Airflow" is completely useless
- Maximum Static Pressure is measured when there is 0 airflow, meaning its situation when you put a fan flat on the table and force it to spin blocking all the air going out. And again, situation like that is completely unrealistic and has no real world use for PC fan, thus Maximum Static Pressure is completely useless.

So, fan manufacturers are trying to lure you in to buy their products giving you useless information that has no impact in real life performance. What do we do now? What to look for?

The P-Q curve. A P-Q curve is a relation between the Static Pressure and Airflow, it describes fan's performance over wide range of situations. It's measured by locking a fan in a fully closed box and measuring generated static pressure and airflow (from outside the box) while slowly opening the exit of said box while the fan is at constant RPM. This link describes the process very well http://www.arx-group.com/pq.html And here is an example P-Q curve:https://martinsliquidlab.files.wordpress.com/2013/02/cougarvsgt15part4.png

As you can see, at the left most side where the curves begin we have the "Maximum Static Pressure" that fan manufacturers love to list on their product pages. For that Pressure airflow equals 0 (test box is closed). After that the box starts being opened so airflow is releases (it increases) but due to bigger opening fan generates less pressure and it goes on like this until airway is fully open and air can flow without restrictions.

Somewhere in the middle of the curve, we have an approximated "real life application", where for example a radiator is blocking about half of the air that blows into it while letting other half through. So what's the big deal? The big deal is that a fan with lower "max static pressure" can perform better than a fan with higher "max static pressure" as the curve is not linear. Just as you can see in attached example, where we come to the "real life" performance area the "weaker" fan shows better characteristic than the fan with bigger numbers.

So how do we get the P-Q curves? We look for independent testers like http://www.coolingtechnique.com/ or we contact manufacturers. I did both, i contacted eight fan companies and here are the results:
Corsair: does not understand what P-Q curve is, silent after escalating
EVGA: no specification on first request, silent after escalating
BitFenix: silent after first request, manual reminder, provided specification on first request
EKWB: provided specification on first request
Noctua: provided specification on first request
Fractal Design: provided specification on first request
NZXT: no specification on first request, successful escalation, specs received
BeQuiet: no specification on first request, failed escalation, refuse to provide specification "we do not give that kind of information to externals"

Some companies i contacted would either share the P-Q curve right away or after asking more persistently. Few companies went silent after simply stating "we do not have the P-Q curve". One company admitted straight that they have the P-Q curve but they are not allowed to share it with customers. The source correspondence is displayed over at my video https://www.youtube.com/watch?v=Ccu7mcJPbb0

The gathered data, both from correspondence with manufactures and graphs available at CoolingTechnique was combined in interactive google spreadsheet:
https://docs.google.com/s...MZ0/edit#gid=443035340

tl;dr
- Do not trust the CFM and mmH2O numbers on manufacturers website (they are useless max values).
- Do not compare fans by their max CFM and max mmH2O (again what manufacturer post).
- Seek reliable fan tests with established testing methodology or contact manufacturers to provide P-Q curves for their fan products.

Thanks for sharing this with the community BR awarded 
 
I've used martins liquid lab for information for years

KeksimusMaximus

bcavnaugh
Wow! And how long has this "P-Q curve" be out in the Public Eye?

The lenght of P-Q is not that much important, as how "long" it is would gives us exactly same information as fan product pages, the lenght of PQ curve is maximum airflow and height is maximum Static Pressure.
 
What you want to look at in the P-Q curve is the airflow and static pressure, somewhere in the middle of the curve, because that's the real life performance of a fan. See the photo here https://martinsliquidlab.files.wordpress.com/2013/02/cougarvsgt15part4.png we discard the left/right hand max specs and instead look at the middle of the curve, and from there we take a note of fan real life airflow and using that middle of the curve airflow value we can somehow compare different fans (as long as different fans also have P-Q curve).

Cool,

Yes - TY KeksimusMaximus for posting this

Thermal Bench is solid, a VERY knowledgeable guy runs it, I know him personally and tries to stay above playing favorites and just giving solid info.
Don't recall seeing him to a P-Q curve but mention it to him, see what he says, might be something he would consider doing.

GGTV-Jon
Any chance of this being put as a sticky or linked too from the already stickied fan topic?
This had fallen back 3 pages and with the default topic day range filter it will not be seen or found with the default range

DONE  
 
https://forums.evga.com/FindPost/2938402

FYI.....
 
https://noctua.at/en/nf-a...-to-nf-f12-and-nf-s12a