Relationship between CFM and RPM

[RyanM]

I was wondering - What's the relationship between CFM, RPM, and dB levels?

Let's take a hypothetical dual-bearing fan. It puts out 100 CFM, 45 dB, and runs at 3000 RPM.

Now, let's cut the RPM in half. Is the relationship linear, and the CFM and dB would be cut in half, resulting in 50 CFM and somewhere around mid-30's dB, or is it exponential, and it would move less than 50 CFM?

Any ideas?

 

[Eskimo]

As far as speed versus noise there has been some emperical studies which have shown a logrithmic relationship, which makes sense since dB is a logarithmic unit of measure.
See the results here.

 

[Mday]

cfm and rpm

there is a relationship. it's highly non linear, and takes some random variables into consideration. so it's more of a probability distribution than anything else.

and it's fan dependent with respect to the motor, and the impaler design (fin shape, number of fins, fin size, impaler depth)

 

[RyanM]

So essentially, there's no general model for figuring dB drop with RPM/CFM drop, given that all other factors remain unstated and equal?

I would've assumed that there was some sort of general rule, which becomes effected by those various design aspects, but still served as a fairly accurate gauge for any fan when precision estimates aren't required.

 

[DivideBYZero]

I have always understood that blade pitch and casing design can have an effect on CFM and noise independant of the RPM.

My 2p.

 

[Mday]

Originally posted by: DivideBYZero
I have always understood that blade pitch and casing design can have an effect on CFM and noise independant of the RPM.

My 2p.

very correct.

 

[dawks]

Yoshis super quiet PC

He used bigger fans (larger diameter) with lower RPMs to move the same amount of air, with less noise produced.. He also made sort of a muffler.. but thats another discussion

 

[RyanM]

Ugh. That's disgusting. He spent all that time sound-proofing, when he probably could've had the exact same effect by removing 2 out of the 3 fans on the power supply. Power supplies do not need anything more than the fan located at the exhaust section - If there is a place to intake the air located elsewhere on the power supply, it'll suck air through there and shoot it out the back.

You don't need to blow air INTO the power supply just to blow it out. It's a small enclosure that's much better sealed than a PC case, so you can rely on a single fan to take care of exhaust duties.

Idiot.

 

[gururu]

for autos:
theoretical cfm (Cubic Feet per Minute of Air Capacity)= (rpm x displacement) / 3456

may apply although I'm not sure what the displacement in cubic inches of a computer/computer fan would be in the equation.

but you can see that the relationship at least in this case is linear

 

[Basse]

Maybe a bit OT but quite interesting...

There is a maximum rpm figure for the fan, when the tips of the fins spin faster that the speed of sound, ~340m/s, they loose the ability to move air. This has to be taken into consideration with the fan diameter/rpm design.

Reg
Tomas

Edit: Maybe some of you mathgurus can calculate what the maximum rpm for a 10cm diameter fan would be? /Edit

 

[DannyBoy]

Originally posted by: MachFive
Ugh. That's disgusting. He spent all that time sound-proofing, when he probably could've had the exact same effect by removing 2 out of the 3 fans on the power supply. Power supplies do not need anything more than the fan located at the exhaust section - If there is a place to intake the air located elsewhere on the power supply, it'll suck air through there and shoot it out the back.

You don't need to blow air INTO the power supply just to blow it out. It's a small enclosure that's much better sealed than a PC case, so you can rely on a single fan to take care of exhaust duties.

Idiot.

Well either way I wouldnt mind considering a project like that.

I would much rather have a silent PC than a hoover going 24/7 in my room.

I managed to take care of my server noise problems in a very technical way.

I modded it into the bottom of my pine wardrobe - EDIT: And i use the waste heat to vent one half of my wardrobe which acts as an airing cupboard

Dan

 

[RyanM]

Hahahahahahah. That's hysterical. 2 birds with one stone is always a good way to go.

Seriously though - It's really easy to make a silent PC without any of that fancy soundproofing. You take some 92 CFM 120mm fans from Enermax, mod your front intake to accept a 120mm fan, cut a blowhole in the top of your PC and make it use the other 120mm fan, then you buy an SLK-800 and mount a 40 CFM 92mm fan.

Then you undervolt them until the whine of your hard drive is louder than the fans. Voila. Throw in an air filter for the front intake, and you're all set.

 

[DannyBoy]

If i lowered them to less noise than my hd i think id probably kill my processor and everything else in the comp :Q

Not a good thing to do when you have a barracuda *Thumb Down*

The barracuda dont make the slightest bit of :music:

I came up with the wardrobe idea by accident btw, perhaps i should market them

Want a self heated airing cupboard / wardrobe / pc all in one? Then try the new Compdrobe

 

[everman]

When you compare the relationship between rpm and db you need to take into account the type of hardware in the fan you're talking about. Some fans are just junk and will make noise no matter what. But if you were to make a fan that lets say used the same time of motor tech used in a seagate cuda V then you could get a rather high RPM with very low noise. (kinda dreaming about the perfect case fan there)

Personally I use a 120mm side fan in my case which is rated at 70cfm max. But I put it on a fan bus and is run silently unless I'm gaming w/ headphones. Greater blade surface area = greater cfm.

 

[RyanM]

Well, even more importantly:

Greater blade surface area = Greater CFM @ Lower RPMs w/ Even more lower noise.
Greater blade surface area = Even more greater CFM @ same RPMs w/ Lower noise.

;-)

 

[Howard]

Originally posted by: gururu
for autos:
theoretical cfm (Cubic Feet per Minute of Air Capacity)= (rpm x displacement) / 3456

may apply although I'm not sure what the displacement in cubic inches of a computer/computer fan would be in the equation.

but you can see that the relationship at least in this case is linear

Real CFM = (RPM x displacement (in^3)) / 3456 * volumetric efficiency