Fan CFM ? in push/pull design

[Anemone]

Ok budding aerospace engineers (it sounds good no?)

In a push pull design do the CFM passed equal the adding of the CFM of both fans, or is there some rough % of each fan that is applicable to arrive at the total CFM that passes through the rad?

Thanks again!

[greenman100]

LOL

I take it you assume the fin thickness/density has no effect, as you didn't give this info

you certainly do not add the free-flow CFM ratings together

think <80% of free flow rating of one fan

probably WAY less

[Zogthetroll]

well, this is a topic that has been thouroghly hashed out a number of times, so there's plenty of material on it around here. If memory serves, (and don't take this as fact, please do some searching to confirm it) the number I remember was that a push/pull config with similar fans netted about a 40% increase over a single fan (pulling).

ha, dang, beat me to it.

[BillA]

data ?
I have tested such and NEVER seen anything like that

DATA guys, keep it real

[pHaestus]

http://phaestus.procooling.com/shrouds-BIX.GIF

nor I

TOO MUCH NOISE for too little cooling gain

[Mars]

40% over a single fan setup? I doubt that. And yeah, I agree. Way too much noise and double the cost.

[Anemone]

Well it wasn't for cooling's sake persay, but rather that DUE to cooling I'm vaguely trying to see if I can balance the airflow before I cut more holes

Thanks for everyone's input though!

[Butcher]

Even were there no restriction you wouldn't get double CFM with push/pull - 2 fans in series are not going to give double airflow.

[Razor6]

Originally Posted by Butcher

Even were there no restriction you wouldn't get double CFM with push/pull - 2 fans in series are not going to give double airflow.

With no restriction and no interference they should get double the static pressure.

[BillA]

Razor6
got data ?
did you look at pH's test results ?
and . . . .

[myv65]

With no restriction and no interference (what's the difference between "restriction" and "interference"?) you will essentially not have any static pressure.

To the original question, this is little different than two centrifugal pumps in series. Much like that, most people who answer have no clue about which they speak. I will give you the "general answer" without specifics as specifics depend on the actual fans, resistance, etc.

When you add a second fan or pump in series the flow rate will go up. Each fan or pump still follows its characteristic curve, however. Each device is passing more flow, but more flow equates to higher losses. Those higher losses are divided across two devices. You need to find the new point on the characteristic curve where the new losses / 2 meets the curve. You can easily determine absolute bounds for this. With no resistance, flow would truly double (not truly possible). With zero flow, static pressure would truly double (no benefit as without flow there is no cooling). In the "real world", I'd hazard that adding a second identical fan would probably bump net air flow to the tune of ~1/4 to 1/3. This is probably (as shown by Phaetus' data) nearly indiscernable to the CPU.

[BillA]

I ran these tests too several years ago with my home-built wind tunnel (thermal anemometer and digital differential manometer) and never saw even 10% increase in flow, none in dissipation (given the substantial 'error' in such measurements)

[jaydee]

I have done a lot of half assed testing on this and have come up with similar results. Louder and the water temp not much cooler. To be honest I don't get much better results using a shroud either but the shroud gave better numbers than adding a second fan.

Of course this is with just cheap DMM's and not great ambient temp control.

[Les]

myv65 is migrant visitor these days.
His words are well chosen and the science is correct
Suggest that, those who have not, read his article

[myv65]

Thanks Les. I'm not sure the "AMDMB" links to the articles always work anymore. If not, users are advised to visit the "new" site Ryan is running at www.pcper.com

[Anemone]

Well thank you all - seems the definite core advice is "don't use push pull designs they're idiotic" is without doubt.

Back to the design board

[Butcher]

Originally Posted by Razor6

With no restriction and no interference they should get double the static pressure.

CFM/airflow and static pressure are two entirely different quantities.

[pHaestus]

Hi Dave
Long time no see!

[Razor6]

Originally Posted by Butcher

CFM/airflow and static pressure are two entirely different quantities.

I didn't say they were the same, I was talking about max pressure when you have no flow. :shrug:

Also when I said no restriction and no interference I was speaking hypothetically. As in, if you could attach two fans in series while not incurring additional restrictions to air flow and could prevent the first fan's turbulence from affecting the second fan.

[Razor6]

Speaking of fans in series Cooler Master has this new dual fan unit the Dual Storm. The first fan looks like a turbine blade and the second fan spins in the oposite direction and has more traditional blades but swept back.

Dual Storm (slow)

Does it really work as advertised? Who knows, but it does look like an interesting design.

[BillA]

I'd have more faith in Delta GUB0812VHG
6 blades on top, 3 below counter-rotating
no specs that I can find, not too noisy

I'll put on a MCW80 and test

[HammerSandwich]

Sounds like the Delta GFBs.

[BillA]

Thanks for the link HS, I was stumped

procooling is a resource

[Tempus]

nice looking fans (took a peak at the 120mm in the spec sheet.)

I'd love to see a review on them.

[BillA]

the noise on those 120s ?
aiiii
I have an 80mm VHE @ 47.5dBA, no whine - just air turbulence