Draw The Air Through The Radiator Or Push It Through

[Methious]

I've heard both, currently I'm drawing the air through my radiator (Swiftech triple 120mm). Which is better?

[flashive]

I'm pushing, but unfortunatly I do not know which is better. But other than saying it 'feels' like a more efficient way, I do not have any grounds to claim that it is any better.

So I'll be pleased to hear which is better as well

[Apoptosis]

I've tried both and didn't notice a big difference. I'm pushing air on it also.

[Methious]

After I get the blocks back on I'll run the temperatures and switch, the triple 120mm swiftech is side mounted fans out for easy access. Huge radiator if I don't like it I can always use it for a club

[Bio-Hazard]

Every radiator type and make performs differently depending on the fan type and CFM rating of said fan.................

Soit's always best to test your system in bothconfigs to see what works out the best for the components you're running.

It used to be the rule that pulling the air through was generally the best solution (less dead space from the center hub, this is also where a fan mounting shroud comes in handy), but with fin count and radiator thickness constantly changing, it's almost impossible to go by any sort of single standard, so test and test again. I'd be willing to bet that the most you're going to see is maybe 1 or 2c differance in temps regardles of which way you have the fans

[kenc51]

I was always told to have the fan(s) in the pull config

[Methious]

I ran them in both push and pull for an hour, pull seems to work better, like Bio said 1-2C difference so I ended up in pull as it ran lower temps.

BTW. Thanks for mentioning the Yate Loon fans Ken I really like them and now I can barely hear the machine sitting right on top of it, and thanks to Bio for pointing out the website I found them on uber cheap (3 43 cfm 120mm $9 bucks total)

[T-Shirt]

if there is some space (enclosed area called a plenum) between the fan and radiator, with enough area for the air pressure/suction to equalize, it shouldn't make a difference. the cooling is -X cfm@ y º over the radiator transfers D BTU. The temprature difference of the air (considered a fluid, for this calculation) entering and leave the radiator is call the Delta.
The factors that MIGHT change the results in a VERY MINOR way are the heat from the fan motor in a push setup (not really a factor), the size/design of the plenum (important in push or pull, air pressure MUST equalize and air MUST flow evenly across the radiator surfaces) AND (the biggie!) the design of the fan (some blade shapes are better at suction)
Overall I always choose pull to make vacuuming out the inflow side of the radiator easy to do (a few blocked fins or even a thin layer of dust effect heat transfer more than all the other factors combined.

[Miazga82]

You actually want see any differences in temperatures between pulling and pushing.
For cleaning the radiator its best to pull. I've set it up for pulling, cuz you don't have to take the fans off to remove the dust and dirt from radiator. And you'll see, after few months there will be lots of dust.

[martini161]

i personally use a push pull, but only because i have mine mounted externally

[markkleb]

if there is some space (enclosed area called a plenum) between the fan and radiator, with enough area for the air pressure/suction to equalize, it shouldn't make a difference. the cooling is -X cfm@ y º over the radiator transfers D BTU. The temprature difference of the air (considered a fluid, for this calculation) entering and leave the radiator is call the Delta.
The factors that MIGHT change the results in a VERY MINOR way are the heat from the fan motor in a push setup (not really a factor), the size/design of the plenum (important in push or pull, air pressure MUST equalize and air MUST flow evenly across the radiator surfaces) AND (the biggie!) the design of the fan (some blade shapes are better at suction)
Overall I always choose pull to make vacuuming out the inflow side of the radiator easy to do (a few blocked fins or even a thin layer of dust effect heat transfer more than all the other factors combined.

wow, great answer.
I went with pulling in my comp using a plentinum as well. I noticed a BIG difference once the box(plentinum) was sealed.

[stev]

i personally use a push pull, but only because i have mine mounted externally

I'll have to agree on this setup if the room allows. A push and pull works the best, but requires higher power requirements.

As Bio-Hazard mentioned, the tighter the air-centers (fins density), the better it is to pull the air.

Pushing air through tight fins causes back pressure to constrain the air flow. This in turn will cause the fans to slow a little or cause the fans to draw more power to try to overcome the resistance. So, pushing is the hardest thing to do.

Pulling air is ideal since the back pressure isn't the problem, however, it too has limits like that of pulling thinks through a straw.

Setting up a slightly faster, more CFM fan to push air in and a moderate fan on the other side to pull air through, the combined flow problems are greatly reduced.

[DMB2000uk]

I totally had to look up the definition of plenum, but it was what I thought, a fan/radiator shroud.

I'd love to have one to maximise cooling, but with my case, I only just managed to cram the fans & radiator in as it is, so there is no chance in this case (which I happen to love to bits).

Dan

[Bio-Hazard]

All depends on the radiator design and fans......;) It's been proven time and time again that for my radiator for example, the thing works best at low CFM/low speed fans with the fans blowing on the fins, as you increase the speed/CFM of the fans, the differance sort of equals its self out..............

As for my rig, I use a shroud designed by the radiator designer/maker with everything all nice and sealed and the fans pulling the fir through............. That seems to work best for the components used in my setup. System idles right at room temp or slightly above with the fans set at 7 volts.

[MakubeX]

It seems most people say that pulling is better.

From my experience, when I had a Swiftech 2x120 rad on the ceiling of my case, I tried both configurations and got the exact same temps.

Now I have two heatercores, one at the back and one on top of my case. One is on pushing and the other on pulling configuration and both work great. Just set it however it's more convenient for you and for the airflow inside your case.

[PCFlip]

ya if you have a good seal around the fan and radiator then pulling and pushing will yield pretty much identical results, although some people will claim that pull is better for a fan then pushing because if you push more air then what can channel through the radiator fins then you could create back pressure, where the air has no where else to go then back through the fan itself and in which would reduce your cooling efficiency, I have seen this happen before but not often, it really depends on what type of fans your using and how much CFMs they are pushing.

[Bio-Hazard]

I think that's what I said above.............

Test and test again to find out what works best for the componnets you have in your system. Just like everything else in the computer world, no 2 systems will perform alike even if they have the same parts.

And push can and often does yeild differant performance from pull due to the fact of the dead air flow area directly under the center hub of the fan............... You can make the area smaller with a shroud, but with pulling air through, you can pretty much get rid of it altogether.

[Methious]

I'm going to stick with pull for a while, when I went to push I could feel air coming back through the fans. I do really like the water cooling now, it's broken in a little and even on the 90nm X2 6400 heat pump I'm seeing 27-28 idle and sub 40 full load for extended periods. If it'll handle that hot chip I know it'll handle any thing else I've been looking at.

[Methious]

Just a minor update, I finally got to the hardware store and replaced the stand off screws I use to distance the radiator from the case. The original bolts were a half inch long (all I had with the same threads at the time), the new ones are 1 3/4" and push the radiator well out from the case, idle temps went from 27 to 23/24C Needless to say 24 cents worth of bolts and now I'm completely happy with the water cooling move!

Thanks for all the advice and patience!

[markkleb]

good job and congrats on the improvement.