Fan Laws

[Shrek]

I had not realized these fan laws

• The 3 Basic Fan Laws - Eldridge (eldridgeusa.com)

If one wants twice the flow rate from the same fan it needs eight times the power; for me this is very motivating to use a larger fan or better heatsink.

 

[freeagent]

f I want twice the CFM from the same fan it needs eight times the power

Oh yeah I have seen 200v 140x38ish cooling components on a CNC Lathe.

 

[Super Firm Tofu]

Oh yeah I have seen 200v 140x38ish cooling components on a CNC Lathe.

And by 'CNC Lathe' he means his PC.

 

[Shrek]

I wonder if there is a similar law for noise production.

 

[freeagent]

And by 'CNC Lathe' he means his PC.

Lol no I run lathes and mills, this one was a Doosan Puma 2600 I believe.. the dual spindle one. I actually just quit that job because feck that guy. Start my new job tomorrow actually.. probably won’t be there long though, we’ll see.

 

[Super Firm Tofu]

I know. Just a failed attempt at poking fun at your slightly above average air flow.

 

[Bill_Bright]

Despite the claims, those are not "laws". They are just formulas used to "estimate" values while out in the field. There are just too many other variables. Air density, for example - which is constantly changing due to temperature and humidity.

for me this is very motivating to use a larger fan

Well, the main advantage to using a larger fan (IMO - because I really hate fan noise) is that you can get the same amount of air flow (CFM) while spinning at a slower RPM. This is assuming identical design, identical quality. The only difference is the length of the blade or diameter of the fan (120mm vs 140mm, for example).

But CFM is based on much more than just size (blade length) or rotation speeds. Blade width and pitch are critical components. As is the design and manufacture of the leading (cutting) edge. A razor sharp, smooth cutting edge is going to slice through the air with much less effort (and noise)than a thick, dull and jagged edge.

Remember, a fan blade is essentially the same as, and uses the same theories and principles as a propellor blade, which uses the same theories and principles as an airplane wing. It is very high-tech.

This means a quality fan must be designed to slice through the air with minimum friction, then scoop up the air and push it out.

One cannot just mold a plastic strip, put a little twist in it, then attach it to a spinning motor hub and voilà! Instant "quality" fan.

I wonder if there is a similar law for noise production.

Since fan noise consists of the noise created by the blades as they "chop" through the air, as well as the noise made by the bearings (which changes with age and wear), and as the moving air molecules are pushed and pulled through (and bang into) case vents/grates, and the fan housing, I don't see how there can be such a "law".

 

[Shihab]

Ah yes! Affinity laws. Easiest hydraulics grades ever!

That said, they have several constraints, chief of which is requirement of similar efficiency and dynamic properties. Comparing any pair of fans/pumps should be based on each fan's set of characteristic curves. A bigger fan may not necessarily be better power-consumption wise than the smaller one.

Despite the claims, those are not "laws". They are just formulas used to "estimate" values while out in the field. There are just too many other variables. Air density, for example - which is constantly changing due to temperature and humidity.

All "laws" of physics, or at least classical physics, describe something under ideal conditions. Fan/pump/affinity laws are no different.

Air density does change, but it's significant in this case. Operation of the fan itself wouldn't change it much (flow is practically incompressible here), and typical temperature variations are too small to yield significant change in density. Between 0~100c, a 20c change in temp would affect air density by, at best, ~5%.
A 20c delta is very extreme...

 

[Bill_Bright]

Air density does change, but it's significant in this case.

I assume you meant "insignificant".

All "laws" of physics, or at least classical physics, describe something under ideal conditions.

Kinda, sorta, but not really.

It is not about being under "ideal conditions". Laws in physics take into account the conditions, whether ideal or not. That is exactly why I said the above are NOT laws - simply because they do not take into account all the variables.

And here, there are just too many variables.

For example, what is the "ideal condition" for a computer case fan? 70°F @ 50% humidity and 0mph wind? Says who? What about barometric pressure? Dust in the air? Input voltage? Blade design? Duct/vent characteristics? Positive/negative case pressure?

Again, those "Laws" Shrek noted are just basic formulas to "estimate" needs for the field engineer.

 

[Deleted member 185158]

One law I follow with my Delta AFB1212SHE fans..... 1.6a with 150 cfm....

Don't get your finger near them.

 

[DeathtoGnomes]

My variables for these fan laws...

Fan Law 1:
CFM = C fan Move
If CFM >= 1 = Fan is on.

Fan Law 2:
SP = Some power = is better than none.
IF SP <= 1 = more fans needed.

Fan Law 3:
HP = Half Power = CFM
IF HP =/= 1 than HP = CFM/2 = less noise.

Rinse/Repeat.

 

[Shihab]

I assume you meant "insignificant".

Yep. My mistake.

Kinda, sorta, but not really.

It is not about being under "ideal conditions". Laws in physics take into account the conditions, whether ideal or not. That is exactly why I said the above are NOT laws - simply because they do not take into account all the variables.

No human law can take into account all conditions. Even Newton's laws of motion, the basis for practically all applied physics, fail once you go extreme enough in velocity or scale.
Laws merely encode observations of a relationship between multiple variables or properties, which can be verified for similar cases under the same conditions of the observation.
A law is not required to cover everything, it only need to be correct and testable to what it specifically observes. Darcy's law is provable for -controlled- porous media flow, but it's useless by itself in modeling -say- contaminated water flow because you also have law of conservation of mass in play. The latter itself fails, in its classical form, to cater for effects such as nuclear decay. Newton's law of viscosity holds for its class of fluids, but try applying it to paint. Gravitation law does work (barring above limitations), but once you're inside an atmosphere, you'll have to subtract drag/buoyancy to get any meaningful force. etc, etc...

I admit that "ideal conditions" was not the best term to use. What I meant that laws show relationship between two or more variables with everything else being equal. Although they may indeed have limited ranges of physical properties within which they hold (referring again to laws of motion).

For example, what is the "ideal condition" for a computer case fan? 70°F @ 50% humidity and 0mph wind? Says who? What about barometric pressure? Dust in the air? Input voltage? Blade design? Duct/vent characteristics? Positive/negative case pressure?

Affinity laws imply that fluid parameters (density, pressure, temps, etc) are the same between the two cases it observes. And this is ok, because those params are independent from the observed ones and vice versa, within the ranges you'd use a pump or a fan on, of course. You can have different discharges at the same temperature and density, you can have different rotational speeds without -significantly- affecting flow temperature, and so on.

Again, those "Laws" Shrek noted are just basic formulas to "estimate" needs for the field engineer.

How a law came to be and how it's used are two different things. I agree that using affinity laws like this gives only rough estimates, but that doesn't deny that it is a sound law.
Engineers are idiots; we will abuse anything to get a quick estimate, even if it wasn't meant to be used that way. Don't hold these poor formulae accountable for our sins...

 

[Shrek]

I have been wondering about the cubed law... but think I have it; double the RPM and that doubles the flow, but the energy goes like v^2 and so one gets the cube.

And the static pressure law, since the force is due to a change in moment and that goes like v, so one power less than the cubed law.

If I get bored, I may try plotting some data to see how well it fits the 'fan laws'
Noctua

 

[Jacky_BEL]

I wonder if there is a similar law for noise production.

Yes there is.

I took note of a formule i found and kept it handy in a txt-file.

relation of fan speed vs noise

noise = speed^5
speed/2 = -15dB

Now, it is not in the correct mathematical notation, but the take away is:

Doubling the fanspeed wil give 15 dB increase in sound and halving the fanspeed wil give a 15 dB noise reduction.

I have not tested this because I don't have the equipment to do so.
A 15 dB increase is a lot , but doubling the fanspeed is also a lot.
The minimum to maximum fanspeed ratio is not a really big value, so it seems it could be a valid relation.

 

[Deleted member 185158]

Or just see what the Spec sheet says

59db max at 1.6a max at 150cfm max.

Wasn't hard, 30 seconds of searching and didn't have to follow the law. In fact, I might have broken some laws during the search process.

 

[Shrek]

I have not tested this because I don't have the equipment to do so.

I hope to use this data https://noctua.at/en/products/fan if I have time

Or just see what the Spec sheet says

59db max at 1.6a max at 150cfm max.

Wasn't hard, 30 seconds of searching and didn't have to follow the law.

I'm after laws not just a single data point

 

[Deleted member 185158]

I hope to use this data https://noctua.at/en/products/fan if I have time



I'm after laws not just a single data point

The beauty part is the data point was used from those "laws" and you are given numbers.

These laws may help if you are building said fan, maybe you have a 3D printer and experimenting.

But otherwise useless to me, you, the next guy. We just buy fans and install em'.

Some need static pressure, some need cfm, some need low noise, all of which is on your data sheet.

Then what happens is a couple guys gonna argue the semantics of wording the ascribed meaning to "laws" and "physics" and who knows what else we'll read about up in here.

My opinion of course, so that doesn't get misinterpreted......

 

[Shrek]

Each to their own

 

[Deleted member 185158]

Each to their own

For certain!

Interesting enough, you can also look up the CFM required to cool certain amounts of wattage.

Like 2U or 4U server cases as an example. This give you an idea of what to shop for!!

Watts per square foot or Kw per square foot would require X cfm but also take into account internal restrictions and so on.

How deep can a fans rabbit hole go? Pretty deep to be technical about it.

 

[Shrek]

One example

I have room fans that I slow down with a variac and wondered about the cooling of the fan motor.

Now I know (thanks to these laws) that halving the speed halves the flow BUT reduces the power by a factor of eight; so, I do not need to concern myself with the fan motor overheating.

For me this is a big deal as I like the air to circulate in the bedroom, but without any distracting noise.

 

[claes]

Yes there is.

I took note of a formule i found and kept it handy in a txt-file.

relation of fan speed vs noise

noise = speed^5
speed/2 = -15dB

Now, it is not in the correct mathematical notation, but the take away is:

Doubling the fanspeed wil give 15 dB increase in sound and halving the fanspeed wil give a 15 dB noise reduction.

I have not tested this because I don't have the equipment to do so.
A 15 dB increase is a lot , but doubling the fanspeed is also a lot.
The minimum to maximum fanspeed ratio is not a really big value, so it seems it could be a valid relation.

Not sure about this one

be quiet! Silent Wings Pro 4 120 mm PWM Fan Review

be quiet! updates its successful Silent Wings 3 lineup of premium PC case and radiator fans with the new Silent Wings 4. Today we examine the Silent Wings Pro 4 120 mm that aims to be a master of all trades, offering multiple RPM operating modes on top of an updated fan design for improved...

www.techpowerup.com

 

[freeagent]

Silence and high performance do not go hand in hand usually.

I love strong fans that kick sand in the face of those other wimpy fans

I think anything that's 100-150cfm at 12v is pretty decent for the hardware we use.

If someone could catch their CPU cooler or video card moving from turbulence I would be uber impressed.

@ShrimpBrime should recognize this sound

A little air flow

Noctua iPPC, Meshify C, Thermalright FC140

youtube.com

 

[Shrek]

For me noise is a BIG issue.

 

[Jacky_BEL]

Not sure about this one

be quiet! Silent Wings Pro 4 120 mm PWM Fan Review

be quiet! updates its successful Silent Wings 3 lineup of premium PC case and radiator fans with the new Silent Wings 4. Today we examine the Silent Wings Pro 4 120 mm that aims to be a master of all trades, offering multiple RPM operating modes on top of an updated fan design for improved...

www.techpowerup.com

If I read the graph correcly , @ 30% PWM the fan does 1000 rpm with 30 dB(A) fan noise.
There is not an exact datapoint at 2000 rpm , it falls between datapoints @ 70% and 80% PWM.
2000 rpm is about 75% PWM , and i read the corresponding noise of 44 dB(A) on the graph.

It is not exactly 15 dB , but I think it is a fair approximation.

 

[Shihab]

If I get bored, I may try plotting some data to see how well it fits the 'fan laws'
Noctua

Wouldn't reach any meaningful conclusions with those tables. The laws don't hold for different fans unless they were of the same design (i.e. same impeller shape and material), only varying in rotational speed or diameter (and quite limited int the case of the latter).

Bored enough to grab an anemometer or breakout some old school pitot tubes/manometers?