Push/Pull Theorycrafting

[Leegit]

So for a solid heatsink with high FPI (15+), what do you think is the best config?

1. 2x High Static Pressure Fans
2. 1x High Static Pressure Fan and 1x High CFM on side towards back of case

Curious about #2. If you are able to push air into the heatsink via high SP fan and then pull it out through the high CFM.

Any thoughts?

 

[EarthDog]

I think you are looking in to it WAY too deep for the negligible difference it will make between CFM and Static Pressure.

 

[Conumdrum]

Solid meaning good?

Don't think you'd see a lot of diff myself, you'll possibly see 1-2c diff. Only a test will tell and the choice of fans is endless so it's impossible to know. And it should be done on a test bench, not in a case for the best repeatable results.

 

[Leegit]

I think you are looking in to it WAY too deep for the negligible difference it will make between CFM and Static Pressure.

Pennies make dollars.


@Conumdrum: Someone needs to do this test otherwise I will when I build my rig. I don't have a testbench, but I guess if I'm the only who is about curious this then that doesn't matter

 

[wagex]

I think you are looking in to it WAY too deep for the negligible difference it will make between CFM and Static Pressure.

+1

 

[bluezero5]

heh guys, This is theorycrafting, let's not say he's overthinking. I will join the fun.

alright, there's a problem with the different CFMs, the lower CFM fan will be DRAGGED by the higher CFM fan, the result will be a potential wear on the slower fan, and can break it.

Do high CFM fans in push/pull will be able to generate increase in static pressure to overcome some of the toughest FPIs. So look for that for better result.

 

[Leegit]

heh guys, This is theorycrafting, let's not say he's overthinking. I will join the fun.

alright, there's a problem with the different CFMs, the lower CFM fan will be DRAGGED by the higher CFM fan, the result will be a potential wear on the slower fan, and can break it.

Do high CFM fans in push/pull will be able to generate increase in static pressure to overcome some of the toughest FPIs. So look for that for better result.

Finally someone will accept my need to optimize everything. I got another month of nothing to do. So I'm gonna plan this build out down to the finest detail.

@bluezero5: Do you think a higher FPI count will mitigate the wearing down of the low CFM fan? My intuition tells me yes but don't know Also, is this caused by the high CFM fan spinning faster or just the CFM difference? Perhaps, you could run the high CFM at a comparable RPM?

So I know we have our own cooling thread regarding fans but I found this thread on OCN:
http://www.overclock.net/t/1274407/...-comprehensive-array-of-tests-and-benchmarks#

Currently mining the data and comparing with recommendations from here

 

[manu2b]

I like the idea, it's interesting...
And we are debating all day long about what's the best cooler, what's the best TIM for a couple of degrees difference, so, this one could bring some fresh air( I love my humour!).

 

[bluezero5]

i thought about this some time ago too. so I don't blame you at all.
you are thinking this right?

<-- High Static pressure fan <-- RAD <--- High CFM fan <---- air sucks in from here.

and that the RAD will reduce the CFM through resistance, so when it reaches the lower RPM fan, it will be at closer CFM of the out fan. Right? In theory, that is true. and in push/pull these fans should have the ability to build on one another.

then I thought.. what if?

<-- High CFM fan <-- RAD <--- High CFM fan <---- air sucks in from here.

would that already neglect the need of a high static pressure fan?
cause both fans in push/pull will already generate high enough static pressure together in series. and I believe I was right, though different is only around 1'C per 240 rad space, having 2 fans of high CFM is already doing all the work it needs. (15-20 FPI Rads.)

I later found, that adding a SHROUD, trumphs all of the above, especially if you are using a fan with a large blind spot. (big center) Adding a shroud between the rad and the fan not only improves air penetration, it also seems to be more positively helping the temperature too.

So don't forget a shroud if that option is feasible in the set up.

 

[Conumdrum]

He's on a heatsink, air cooling, not a rad but same principal.

 

[Wipeout]

With good air flow, and a push and pull set, you should be fine.Some people might view this different, but my temps convince me otherwise.I use 8 fans, and they seem to work well in conjunction.

I use a Hyper 212 Evo with 2 fans for push and pull.This set up works really good in both of my cases.On a side note, I used to use covers on my pci slots, but its much better leaving the extras off that are not used.Its really keeps dust down to a minimum.I have really good air flow from front to the back, and added another 140mm fan behind the hdd's. I have strong air circulation, and my temps are great.The inside fan does a nice job keeping video card cool, and stronger stream of cool air flowing.

This is a Rosewill Black Hawk Mid Tower.Its a great case for cooling.I have an Antec DF 85 full tower, and this case cools just as good, if not better.I'm a gamer, and a sucker for full towers, so considering the Black Hawk Ultra for my next case.Air flow would never be a problem.Add a push and pull set up, and watch your temps go way down.One feature this has is a fan hub.Its located near the upper portion of the 5 1\4 bays, a on right side of the case. 2 more fans mounted on inside, attached to the hdd cage.Nice touch.Amazon = 30 dollars cheaper than the Egg.I hate the Egg anymore.

Rosewill Gaming Super Tower Case for Computer BLACKHAWK-ULTRA

http://www.amazon.com/Rosewill-Gami...369548521&sr=8-4&keywords=rosewill+full+tower

 

[bluezero5]

looks nice !! take 'sharper' pictures!

very nice build.

 

[Wipeout]

looks nice !! take 'sharper' pictures!

very nice build.

Thanks.

I have to learn this new camera.I guess zoom vs auto focus should work better.I have always went with name brand on cases, but Rosewill was a pleasant surprise.To bad the fans are all molex connections . It does come with 5 stock fans, so no complaints.I replaced the front and back fans with Antec Tri-cools, and then used the front stock fans on the top.I moved the stock 140mm on the top behind hdd's, and bought another cooler master fan for push and pull set up.Funny, those stock fans are sleeve bearing fans, and I am using both on the top of the case.Bad choice for sleeve bearing fans in a horizontal postion, but there running very quiet for 5 months now.They also push alot of air.Surprising to say the least.

 

[Leegit]

i thought about this some time ago too. so I don't blame you at all.
you are thinking this right?

<-- High Static pressure fan <-- RAD <--- High CFM fan <---- air sucks in from here.

and that the RAD will reduce the CFM through resistance, so when it reaches the lower RPM fan, it will be at closer CFM of the out fan. Right? In theory, that is true. and in push/pull these fans should have the ability to build on one another.

then I thought.. what if?

<-- High CFM fan <-- RAD <--- High CFM fan <---- air sucks in from here.

would that already neglect the need of a high static pressure fan?
cause both fans in push/pull will already generate high enough static pressure together in series. and I believe I was right, though different is only around 1'C per 240 rad space, having 2 fans of high CFM is already doing all the work it needs. (15-20 FPI Rads.)

I later found, that adding a SHROUD, trumphs all of the above, especially if you are using a fan with a large blind spot. (big center) Adding a shroud between the rad and the fan not only improves air penetration, it also seems to be more positively helping the temperature too.

So don't forget a shroud if that option is feasible in the set up.

Actually I was thinking:

Back of Case <--- High CFM <--- Heatsink <--- High SP <--- Middle of Case


with the idea being to inject air into the heatsink with the high static pressure and then drawing the air out with a high CFM. Since the air isn't being blowen into the high SP fan, there shouldn't really be a dragging effect. Perhaps even matching their RPMs would help prevent the fan degradation.

I'm really curious about the role a high RPI heatsink has on the CFM fan when its place in front of the fan.


Did you buy shrouds or make them yourself? What was the clearance for them?

 

[bluezero5]

i make my own.

get an old fan, clip off the 4 beams, and you will just have the shroud left.

 

[Wipeout]

I'm not a theorist, but I know my pull fan runs at 1920 rpm, and the push fan runs about 100 rpm slower.Both fans run on full rpm's with all the green stuff off. In my case, it would defeat the purpose to run fans slower, concerning overclocking, but I would do that anyway.Even at stock clocks.The cooler the better imn a logical standpoint, the push fan is getting more wear and tear,but its also mounted on the heatsink for direct cooling.I have used a Hyper 212 in a all my builds, with cooler master fans, and those fans hang tough.I never had one go bad.I used to run my pc 24/7 365, so those are real world results.What's a fan, 15 dollars.Say, it did crap out in a year, which it didn't. I would feel I got my moneys worth.The only reason I switch out to new fans was to install a new cpu,and replace the heatsink. Moved from a Plus to a Evo.With a good air current, and no restrictions between heatsink and rear fan, air is pulled out the back of the case efficiently.Those temps listed in my specs are a Prime 95, 2 hour test.No need to stress it 24 hours.Gaming will never put that kind of a load on all cores, in any situation.

 

[noname2020x]

One of these days I hope Aero Engineers team up and do a full blown study on push and pull configurations and push or pull configurations. We could all finally put all this theorizing to bed.

 

[Conumdrum]

They have. The answer is 42.

 

[manu2b]

 

[noname2020x]

42 it is.

http://www.amazon.com/Hitchhikers-Guide-Galaxy-Trilogy-Hikers/dp/0330492047