cfm and actual cooling ability

[dxiw]

A general trend is to get a larger fan and then rheostat it down to acceptable noise levels and it still blows more air than a smaller fan. However, I have been thinking about this, and really a larger fan, when run at slow rpms to blow about the same as a smaller fan at max rpms, puts out the air over more area. So when paired with a heatsink there is less "pressure" (not exactly, can't think of a better word) on the metal fins. And if there is less force or rather pressure of the air blowing on the fins its possible that maybe it wouldn't cool as well. maybe instead of looking at cfm we should starting looking at a cfm/size ratio.

 

[Krusty]

The larger fins on the bigger fans do indeed lead to the air flowing over a larger area and often result in the airflow being less direct when attempting to funnel the air into the heatsink. There is, however, a way around this. Flip the fan to reverse the direction of the air and then duct it to suck air through a smaller opening (attach to the heatsink?) and blow the air through the full area of the fan. This way, the fan will still be blowing the same cfms that you want, but the airflow will be being sucked into the fan through a smaller area. This would result, of course, in faster airflow through the heatsink.

I personally do not adjust the fan speed of my heatsink fan and I would not want to use a larger fan on my heatsink and drop the voltage to reduce the noise. The ducting would prove to be more of an annoyance than I can handle. Instead, I use 172MM fans for my case and have them running at 7v instead of 12v. They only push about 100cfm instead of 235, but that's all I really need.

 

[Greedy Guido]

Usually, forcing an axial flow fan to suck drops performance and increases noise. Axial flow fans are not designed to suck.

What you need is a shroud which is at least one fan diameter long.

If you want to get really technical then ensure the any transition has an angle of 15° or less. This is what large commerial fans aim at to reduce losses and noise.

 

[Krusty]

Usually, forcing an axial flow fan to suck drops performance and increases noise. Axial flow fans are not designed to suck.

No, I didn't mean reverse the actual rotation of the fan. I meant flip it around (and shroud it) so that the air flowing through the heatsink is directed by the fan pulling air through the heatsink, through the duct, and out the fan. No fan modification necessary.

 

[dxiw]

i don't think that would still be good enough, i think some simple math can solve this, though we cwould need experimental data. any of you have a 120 and an 80 and a converter duct?

 

[Greedy Guido]

No, I didn't mean reverse the actual rotation of the fan. I meant flip it around (and shroud it) so that the air flowing through the heatsink is directed by the fan pulling air through the heatsink, through the duct, and out the fan. No fan modification necessary.

Thats what I meant. Either way the fan is sucking.

 

[Since87]

Darn it.

I just asked this question with regard to shrouded radiators at procooling and was told matter of factly that the fan should suck.

It makes sense to me that having a restricted inlet to the fan would be a bad thing.

I don't understand why there would be more noise though. Turbulent intake vs laminar intake? Or is this more of an issue when the fan is right on top of a heatsink?

 

[dxiw]

the mroe then fan ahas to work its going to make more noise. If the air flow is restricted its gonna be running higher rpms to force out the same air (if it can) and be louder. The air restriction ahs a lot to do with noise. turn on a fan and hold it, then put it on the table, then restrict it with a net or something. You will see that when you restrict the intake it gets louder and louder

 

[Intrepid]

I did an experiment with an overhead projector fan, and found that it was silent when run with nothing around it, i put some cereal boxes around it like the sleeve of a regular DC fan, and noise went up dramatically.

My conclusion from this is that noise is always caused by air rubbing against surfaces (not the motor or fan blades)

I also note that the size of a fan increases the amount of air moved, not the speed at which it is moved, and that is controlled by RPMs. So in total going from 80mm too 120mm gives a change in M in the unit system CF/M, the bigger the fan the small M gets. Noise, because it is the act of air rubbing against the inside of the case, is IMHO CF/M where CF is constant, because we are judging the surfaces in the whole case, the area of which does not change.

Thus an increase in speed without changing the volume will result in greater noise.

About having a big fan on a small HS, remember in the end the same CF is going through the HS, so all that will change is M, and if you duct a big fan onto a small HS, than M will go back down (and produce more noise), to give you almost the same results as a normal sized fan. Conclusion: get a bigger HS!

I know much of this has been said before but I just wanted to get my theories out there for those who want to pick away at em.

 

[stompah]

Intrepid6546, did you take into account the air hitting the sides and bouncing back onto the blades? That noise u likely hear is buffeting but I wont rule out your theory 100% as the lip of the cardboard could also produce turbulence and noise.

An unshrouded fan will act as a centrifuge and expell most of its air from the sides. Not a perfect centrifuge but one none the less.


With HS cooling you must take many factors into acount. One of which is important to both flow and noise is the center spot of the fan. That is essentially a dead spot. If you build a shroud that reduces diameter too quickly you will bounce the air onto itself and into the dead spot. Not productive for pushing air. Especially if you look at the center section of a 120mm fan over a 60/80mm HS. The center section nearly covers the entire area on the 60.

To properly design a fan shroud will take lots of R&D to get the max out of the fan and HS. Obviously higher blade speed increases air blown off the sides.

Its better to find a slightly larger fan like a 92mm fan on a 80mm HS (or 80 on 60). that may reduce some of the affect of a poorly designed shroud.

And off the shelf shrouds are likely poorly designed and may not work with the many pitches of the fan blades. And running the fan at a lower speed also changes the angle a which the air exits the fan. So if a fan was matched to a 7k rpm fan it may not be as effective at 3k rpm.


But in the real world you will likely see a greater decrease in noise than change in temps. If you speed the fan properly.

 

[Tygur]

I have a few things to say.

First off, it doesn't really matter whether the fan is sucking air away from the heatsink or blowing it onto it. The fan still has to struggle to get that air through the fins. The only way it can really make a difference is if the heatsink somehow offers more resistance when the air is going in one direction than the other.

the mroe then fan ahas to work its going to make more noise. If the air flow is restricted its gonna be running higher rpms to force out the same air (if it can) and be louder. The air restriction ahs a lot to do with noise. turn on a fan and hold it, then put it on the table, then restrict it with a net or something. You will see that when you restrict the intake it gets louder and louder

If the fan has to work harder, the rpm's will actually go down, not up, though the noise level will still rise.

I did an experiment with an overhead projector fan, and found that it was silent when run with nothing around it, i put some cereal boxes around it like the sleeve of a regular DC fan, and noise went up dramatically.

My conclusion from this is that noise is always caused by air rubbing against surfaces (not the motor or fan blades)

The motor has to be making some noise, though it may not be much. As the fan spins faster, the motor would presumably make more noise. Still, I'm not sure how significant it is. It's different from fan to fan. It's the same story with the fan blades. The noise from them is there, and it's a different amount from fan to fan. To conclude that the fan blades and motor never make a significant amount of noise, regardless of the fan is wrong.

I think the fan itself might actually make more noise if you restrict the airflow, as you did in your test, due to extra turbulence around the blades. Basically, while your test did prove that there is more noise if the airflow is restricted, it didn't prove where that noise actually comes from.

Just something to think about. Let me know if my reasoning is flawed.

 

[Intrepid]

Well. I still hold to one truth, a fan pushing air through a case will make more noise than a fan with no shrouds or case to push air through, for whatever reason. To be quite honest I am not the greatest with all of the 'in theory stuff' because I prefer to actually test things in real life situations. And just to clarify I AM NOT saying theories are bad, I like them plenty. Ok back to business, I think there is some noise made by the blades and motor, but it seems to be negligible. As far as air bouncing back into the blades on some fans I know this to be a problem, my current fan on my heat sink (actually forgot what it’s called but it is the kind with all the pins) has a ton of I guess what would be called backflow. Not a good thing IMHO. I wonder if by having two equal fans pushing air into that heat sink, the power of both fans pushing air would eliminate some of the backflow by forcing more air through the heat sink.

 

[Paxmax]

If the fan has to work harder, the rpm's will actually go down, not up, though the noise level will still rise.

I think the fan itself might actually make more noise if you restrict the airflow, as you did in your test, due to extra turbulence around the blades.

If you restrict the inlet of an ordinary axial fan, the rpms usually go up at some point. You can clearly hear the fan rev up, and it also "seems" to make more noise. Maybe it's just the higher rpms thats makes it more irritating. Also You might get higher air velocity at the restricted area that will contribute to the more irritating noise. a wooshing sound?
Anyway Tygur, just break out a fan and rev it up, cup your hand above it and test it yourself.... the fan revs up, period.

 

[Tygur]

If you restrict the inlet of an ordinary axial fan, the rpms usually go up at some point. You can clearly hear the fan rev up, and it also "seems" to make more noise. Maybe it's just the higher rpms thats makes it more irritating. Also You might get higher air velocity at the restricted area that will contribute to the more irritating noise. a wooshing sound?
Anyway Tygur, just break out a fan and rev it up, cup your hand above it and test it yourself.... the fan revs up, period.

Sure, it would make more noise, but how would the RPM's go up? I might actually try something like this, except I'd connect the rpm sensor wire to my mobo so I can check that. I just don't want to open up my case. From what I see, though, what would make the blades spin faster? The only two ways I can come up with is if the air resistance goes down, or if the electricity supplied increases. And neither is happening here, I believe.

 

[pHaestus]

P-Q curves for a restriction on intake vs. exhaust side here:

http://www.nmbtech.com/html/fans/engineering/engineering_101.html#ducting

Take home message is that the size of duct changes for intake vs exhaust side to minimize losses.

 

[MILLTHERM]

Fan speed increases.

Fan usually slow down as you move from freeflow to about 65% of freeflow at which point you get "fan slip" or fan stall. the fan to put it loosely is not "grabbing" the air and becomes less loaded and speeds up. The added noise is both aerodynamic and mech. In fan application your generally most eff. point is 60-80% of freeflow for noise and mech.

Noise is a signature of ineff. The more noise, usually the less eff. the device.

 

[Samoyed]

I just ran a test. I have a duct running from my HSF through the side of my case that sucks in room temp air. When I cover this opening, the RPMs of the fan INCREASE from 5000 to about 5600 according to the software RPM monitor. The fan sounds louder and faster.

I believe that it is "fan slip" as someone mentioned earlier. The fan goes from a mode where it is moving air past the blades in one direction, to another mode where it is moving the same air in a vortex. It is moving the same air because the intake is blocked.

Think of a situation where you were cranking a propellor to move water through a pipe. If the water was blocked from the intake or outflow, the water would being to swirl in the pipe and cranking it would actually become easier as the water simply spun with the prop blades.

I hope this makes sense. lol

 

[woods]

Tygur, the reason the fan blades will spin faster when the inlet is restricted is due to a decrease in drag over the surface of the blades. With less air restricting the movement of the blades, the fan cann move faster with less resistence. It is like on a cars drag. Apply the same power to two objects. One with little drag and the other with more. The one with the less drag will go faster than the one with more while using the same amout of energy.