Push or Pull??

Monkeyspaz5000

Monkeyspaz5000

Monkeyspaz5000

Associate
Joined
1 Jan 2012
Posts
1,946
Location
Scottish Borders - UK
Assuming you can't have Push/Pull setup for one reason or another, what would be the next best?

Push or Pull??

Pushing air through the rad or pulling it?? Or not really a difference at all??

cheers
 
andybird123 said:
Push, but only by about 1c and push&pull together about another 1-2c
Click to expand...

Fan design presumes that on the inlet side of the fan is a volume of air that is available to be pulled through.

When a fan is put in "Pull" configuration, that source of air is constrained by that which has "flowed" through the radiator and you end up with a lower inlet pressure as a result of the inlet to the fan trying to pull air trough the radiator just for the fan to do it's job.

When in push, the EXIT of the fan is slightly obscured by the radiator, but the source of air is not, therefore more air flows...

The hows and whys are widely discussed, but the facts remain measurable and real - push is better - push pull is best because it moves the air away from the obstacle (rad) allowing a greater net flow.

For some experimental data which I think sums it up quite nicely, see

http://www.overclock.net/t/1292456/1-fan-vs-2-chapter-2-push-vs-pull
 
Tattysnuc said:
Fan design presumes that on the inlet side of the fan is a volume of air that is available to be pulled through.

When a fan is put in "Pull" configuration, that source of air is constrained by that which has "flowed" through the radiator and you end up with a lower inlet pressure as a result of the inlet to the fan trying to pull air trough the radiator just for the fan to do it's job.

When in push, the EXIT of the fan is slightly obscured by the radiator, but the source of air is not, therefore more air flows...

The hows and whys are widely discussed, but the facts remain measurable and real - push is better - push pull is best because it moves the air away from the obstacle (rad) allowing a greater net flow.

For some experimental data which I think sums it up quite nicely, see

http://www.overclock.net/t/1292456/1-fan-vs-2-chapter-2-push-vs-pull
Click to expand...

The question is do you have to have the same fans. I understand GT's are among the best pushes but a pretty bad pullers. when i had them in push pull config noise was the main issue but the drop was genuinely negligible.
 
TwsT said:
The question is do you have to have the same fans. I understand GT's are among the best pushes but a pretty bad pullers. when i had them in push pull config noise was the main issue but the drop was genuinely negligible.
Click to expand...

What fans would be good for Pull only?
 
wissel said:
What fans would be good for Pull only?
Click to expand...

I don't know I'm sure someone on here does. We have a collossal range of fans and many of them shift some serious air but not that many are quiet about it. Especially in my Aluminium Lian Li case. I've always been a fan of viper fans but they are yellow.
 
you should match the fans, having mismatched fans means that they will not be cooperating as they should and the slower of the 2 will act as a restriction instead of helping...

push-pull will always only give you marginal gains over push only... if you have big enough rads it will be even more marginal - the only time it gets you more than 2C is if you have small rads and need every ounce of extra cooling to keep things under control

I certainly would never advocate push pull with GT's as the cost is shocking - but by using much cheaper fans you could get the same performance of a GT for the same noise level and far cheaper cost (or just use an over-sized rad with loads of cheaper fans in push)
 
andybird123 said:
you should match the fans, having mismatched fans means that they will not be cooperating as they should and the slower of the 2 will act as a restriction instead of helping...

push-pull will always only give you marginal gains over push only... if you have big enough rads it will be even more marginal - the only time it gets you more than 2C is if you have small rads and need every ounce of extra cooling to keep things under control

I certainly would never advocate push pull with GT's as the cost is shocking - but by using much cheaper fans you could get the same performance of a GT for the same noise level and far cheaper cost (or just use an over-sized rad with loads of cheaper fans in push)
Click to expand...

I hear that on GT's I got a few spares lying around :o The Aero Cool Sharks are fantastic tho
 
ubersonic said:
One used with a shroud (so the is a gap between the fan and the rad).
Click to expand...

if that 0.3C is really important to you then yes, but you'll get 1.5C by switching to push with no shroud instead at the same RPM of fan

but then looking at the link above - for high RPM fans pull only seems to be better than push only
 
Last edited:
Im glad someone asked this question as i was about to ask myself

Would fin count and the thickness of the rad make a difference?
 
mp101 said:
Im glad someone asked this question as i was about to ask myself

Would fin count and the thickness of the rad make a difference?
Click to expand...

higher FPI count tends to mean higher speed fans work best (and vice versa)... I can see low FPI rads working better with pull than high FPI rads... but I'm not sure a low FPI rad would be better on pull than push

^ as mentioned, try it, you might like it ;)
 
TwsT said:
The question is do you have to have the same fans. I understand GT's are among the best pushes but a pretty bad pullers. when i had them in push pull config noise was the main issue but the drop was genuinely negligible.
Click to expand...

That does not make sense aerodynamically. A fan pushing has the same characteristics as a fan pulling. The difference is the availability of fluid to flow through said fan.

Can you substantiate your statement as I did?

As I'd expect, I'm struggling to find fans that work better in pull than they do in a push config. I'm more than willing to concede that 15 years is a long time since my aeronautics and fluid dynamics courses at Uni, and therefore I could well be wrong about this, but I'd like something to read into about it if that's the case please....:)
 
I've just completed my build.

A 240 rad with fans on pull and a 120rad with fans on push pull. All the same Corsair SP fans.

If I have the push pull fans on a high speed they create a horrible noise. If I have the push fans on the larger rad on high speed all is fine.

If I unpug either fan on the push pull then the horrible noise goes away.

It seems like somehow the fans don't like the push pull config that well?!
 
qbazdz said:
Looks like Martin confirmed those findings.
http://martinsliquidlab.i4memory.com/Radiator-Fan-Orientation-And-Shroud-Testing-Review.html

For low heat Pull is better, while push gains advantage with higher loads and more heat.
Click to expand...

That review supports what we've been saying all along...- if we look at the graph:

FanShroudTest1-38mm2000.png



I refer to both points on the graphs without shrouds.
The temp out for Push shows 9.15, whereas the the corresponding PULL for the same fan has 10.10

This whole experiment by Martinslab seems to support the logic I employed, but have not been able to properly explain. :)

The use of 2 fans, one in push and one in pull is far and away the superior choice

FanShroudTest3-38mm2000.png
 
And now if I read the rest of the article, a bit more makes sense.

In laminar flow, ie low fan speeds, PULL seems to be the most efficient performer, in high speed turbulent fans, pull fans are starved of air to pump more by having the radiator obstructing flow to the fan, and therefore the PUSH fan is more efficient.

The delta for low speed fans advantage on low speed is

FanShroudTest4-25mm1350.png


...only 11.91-11.65 = .34C, whereas at high speed the advantage of HIGH SPEED is 10.1-9.2 = 0.90C

Therefore at low fan speeds, the radiator has less impact, and pull is MARGINALLY better, however, as airflow increases, and the amount of heat transferred away from the radiator by the fans increases, the fans in pull become less and less efficient due to the disturbance and obstruction in feeding airflow.

In an Apples-Apples comparison, the PUSH fan orientation performs 2.65x better (0.9/0.34) at the optimum fan speeds.
 
coughs said:
What do you guys recommend on these rads:

http://www.overclockers.co.uk/showproduct.php?prodid=WC-053-XS&groupid=962&catid=1523&subcat=1524

http://www.overclockers.co.uk/showproduct.php?prodid=WC-047-XS&groupid=962&catid=1523&subcat=1525

I have Scythe Gentle Typhoon 120mm 1850 RPM fans? Are there any better fans out there? I am looking for the best performance but hopefully it will be quiet
Click to expand...

GT Typhoons 1850's are the rolls royce, but if you want to work on a budget I've found that the Scythe Jyuni's are great, but when used in conjunction with a fan controller to balance noise/performance.
 
You must log in or register to reply here.
Back
Top Bottom