Push/Pull fans

[Kantastic]

I was reading reviews about fans in push/pull on heatsinks and radiators and I realized something... most people use two identical fans or two fans with similar CFM ratings to use for push/pull, but I don't think this would yield optimal results. The reason being a fan loses a lot of CFM if it's being blocked by so many fins on a heatsink/rad... so wouldn't it be better to use a high CFM fan for push and a lower CFM fan for pull?

 

[Tom20]

I've always thought it's best to have high CFM fans push and pull

My fans on my radiator push 77cfm. I use Yate Loon high speed fans. I am thinking of 4 more for pull.

 

[Binge]

I was reading reviews about fans in push/pull on heatsinks and radiators and I realized something... most people use two identical fans or two fans with similar CFM ratings to use for push/pull, but I don't think this would yield optimal results. The reason being a fan loses a lot of CFM if it's being blocked by so many fins on a heatsink/rad... so wouldn't it be better to use a high CFM fan for push and a lower CFM fan for pull?

Short answer 'No.'

 

[Kantastic]

I've always thought it's best to have high CFM fans push and pull

My fans on my radiator push 77cfm. I use Yate Loon high speed fans. I am thinking of 4 more for pull.

But what exactly will you be pulling? Of the 77 cfm that your Yates are rated to push, only half of that is going through the rad.

 

[BraveSoul]

interesting, i would just make them equal, but my case has more pull then push due to thick dust filters on all 4 120mm fans,, if i open a hole in the case i feel air rushing in so exhaust fans help intake fans

 

[Binge]

pew pew pew

Was originally going to just test the Ultra Kaze 3000 and 3 other fans in the various combinations, but after seeing the results, I added the S-Flex to the mix to get a better feel for what happens between 1000RPM and 2000RPM.

Here's the abbreviations I'm using in this post:
-l = pull
-h = push
UK# = Ultra Kaze #000RPM
U1C = Panaflo U1C
SFx = S-Flex x (E,F,G)

So: Ultra Kaze 3000 in pull = UK3-l
S-Flex G in push - SFG-h

UK3-l = 97.4CFM (percentage improvement over single fan - 0%)
UK3-h = 109.5CFM (0%)

U1C tests
UK3-l + U1C-h = 136.7CFM (28.8%)
U1C-l + UK3-h = 138.9CFM (21.2%)

UK2 tests
UK3-l + UK2-h = 112.7CFM (13.6%)
UK2-l + UK3-h = 123.9CFM (11.6%)

UK1 tests
UK3-l + UK1-h = 97.1CFM (-.3%)
UK1-l + UK3-h = 110.3CFM (.7%)


The following tests are with the UK3 in PULL always (some data is repeated from above, btw).

UK3-l = 97.4CFM (0%)
UK3-l + UK1-h = 97.1CFM (-.3%)
UK3-l + SFE-h = 100.7CFM (3.3%)
UK3-l + SFF-h = 105.6CFM (7.8%)
UK3-l + SFG-h = 109.9CFM (11.4%)
UK3-l + UK2-h = 112.7CFM (13.6%)
UK3-l + U1C-h = 136.7CFM (28.8%)

What's so interesting, IMO....is that everything up to the SFG/UK2 is basically 'free' in terms of noise. That is, the UK3 is so loud that adding even the UK2 didn't add noise (okay, it added a LITTLE--the SFG added basically none though), but added up to ~13.5% extra airflow.

I have no clue what this will do long term to a fan...I'm not sure I'd do it in my own system, but that's because I also won't use a 3000RPM fan in my system

I forget where I posted this before, but pushing fans suffer from back-draft, and pull fans suffer from a lack of air volume. The pulling fans eliminate the back-draft of the push fans and that's the increase in performance. Your pull fan if more powerful than the push fan will yield the greatest results.