Running push+pull for silence?

[Odai]

Hello,

As I've spammed in several of my other threads recently, I'm currently designing an external watercooling box to house pumps, radiators, and reservoirs. I'm trying to find out exactly what I could need in there to determine the best overall dimensions.

One of the issues is whether I run push/pull or push and pull fans on the radiators. I'm trying to gear the whole thing towards silence (or as near as possible to it), and figured the best thing to do would be just to run one set of fans in pull on the radiators (going to be using 2 420mm radiators, 1 in each loop). However, would it be better in fact, given I want to optimise for silence over absolute cooling power, to use 6 fans in push-pull and run them slightly slower? Would that achieve better temperatures for a given overall noise level?

There are plenty of tests out there quantifying the improvement in absolute temperature deltas for using push-pull over push or pull only, but nothing (that I can find) which quantifies the improvement (or reduction) in cooling power for a given noise level on a comparative basis.

Does anyone have any experience on this? To give you an idea, I'd be looking at running a typical fan at less than 1000RPM even under load conditions, and will either be using the Alphacool UT60 or the Monsta radiator models.

Thanks (again)!

[rge]

I run push/pull at 700 rpms, because that is my definition of quiet, and 700 rpm still drowns at my D5's on setting 4, so I can just barely hear air movement nothing else. When I just run push or pull only, I get ~2C higher water temps (measuring ambient and water with calibrated probes), at the same noise level. And my rad is a thinner XSPC rad.

With thicker rads and similar low fan speeds (700 to 1000 rpm range) you should get at least similar improvement again without increasing noise, but that will be based on heat load etc.

However at higher fan speeds like 1400 rpms or so, you may get only 1C better temps with push/pull but at higher noise than push or pull. At lower fan speeds push/pull doesnt increase noise appreciably.

[Martinm210]

I did run one test a while back. Theoretically doubling the fans adds 3dbA, and I have heard some folks say an RPM increase will do the same, but I don't agree. I have seen push/pull GTs pull a better noise/dbA result than one in push on my fan test rig. I have also seen some fans like the swiftech helix eliminate harmonics issues by push push vs just push or pull. It seems like blade stability issues often don't show up until 1400 and higher on some fans.

It does depend on the actual fans but generally, I have only found good from pish/pull except for the price of more fans.

I can't find that old test, but I'm convinced push/pull is better on noise/performance in general.