Just how quiet is water cooling compared to air cooling?

Thanks everyone for the replies. It's kind of confirmation I needed. :)


JeffyB said:
At 600rpm they are inaudiable, yet the water loop will be removing much more heat as water is a better conductor and the radiator has amuch larger surface area than any of those heatsinks.
Click to expand...


I know the science behind the conduction and convection. There's more to it than just the surface area ;)

Water is fine. But the question is will I test ethylene glycol and R-113 in the water cooling loop as the working fluids! :eek:
 
Last edited:
Ice Tea said:
Watercooling being quieter than air cooling used to apply to the old days when air coolers used high speed 80mm and 92mm fans but a single 120mm / 140mm low rpm fan on a pipe cooler will always be quieter than 3 low rpm fans on a radiator plus the pump noise for the sake of a extra 200mghz overclock.

And anybody that thinks a computer running with temps in the high 30s is better than a computer with temps in the 50s is suffering from OCD as all you have is a lower temp reading in task bar.

:D :p :)
Click to expand...

bah, depends what you want to run and how quiet. If you're really ocd you could have amassivelhy overclocked cpu and gpu cooled by a passive watercooling loop and have the pump on a fan controller so the system is incredibly quiet and still poweful... And that would be far better cooled annd quieter than the cpu plus gpu fans. Besides which, don't you like the reserator? ;):D:P

Ricochet J said:
Thanks everyone for the replies. It's kind of confirmation I needed. :)

I know the science behind the conduction and convection. There's more to it than just the surface area ;)

Water is fine. But the question is will I test ethylene glycol and R-113 in the water cooling loop as the working fluids! :eek:
Click to expand...

why would you want t unless running sub-zero? - they're worse coolants.
 
Depending on how well they do in reviews i may have New Fanless love with the latest Thermalright sintered heatpipe Passive coolers . :D

Thermalright Sentinal - Thermalright HR-02

 
From my investigations in the lab, and from investigations from other instituitions (I'd be happy to link to these, however they're not free on the science direct website), glycol and R-113 do provide pretty decent vapour side temperature differences via coolant conduction. This was for vapour side mind, however I'm confident that it'll be same for coolant side.

In any case, I'm more than happy to flush the system if I don't get the results I'm hoping for.


On another note I'm seriously tempted to design and manufacture my own rad...
 
Last edited:
Ricochet J said:
From my investigations in the lab, and from investigations from other instituitions (I'd be happy to link to these, however they're not free on the science direct website), glycol and R-113 do provide pretty decent vapour side temperature differences via coolant conduction. This was for vapour side mind, however I'm confident that it'll be same for coolant side.
Click to expand...

A glycol/water mix will perform worse than pure water, how much depends on the ratio of glycol:water, chances are it won't make a huge difference to the end temps of your CPU/GPU.

Glycol only really needs to be used in loops that have both alu and copper in as a corrosion inhibitor. But nowadays everything is copper/brass so straight distilled/deionized with something to stop algae from growing (PT Nuke or some silver) would be the cheaper/easier option in the long run.

Monkey Puzzle said:
Do it! What kind of thing did you have in mind? Would be interested in seeing a deep, widely spaced fin radiator with lots of louvres on the fins.
Click to expand...

Creating a really deep radiator would give a very minimal performance increase, since air has a very low specific heat capacity, so the air is almost completely 'saturated' with heat by the time it has passed through a single thickness radiator.

This is why double thickness radiators (such as the PA120.X rads) don't give a huge performance increase over single thickness radiators, since the temperature delta between the air and the fins of the radiator has become pretty small already by the time the air has passed through a single layer of radiator, and as the delta between the two gets smaller the slower the rate of heat transfer between the two.

To put it simply the only way to make a noticeably better performing radiator than those already in production would be to increase the frontal surface area of the radiator, and then you start to run into size limitations etc.
 
I've heard the idea of air being saturated before - but if that were the case surelyy the air exiting the radiator would be at the samp temp as the water exiting the rad, and it isn't iirc. The pa series radiators and similar deep widefinned ones don't perform above narrow-finned ones because there's a comparable or less surface area.

The reason why i thought deeper radiators didn't do so well was that after the leading edge of the fins (where air first meets the fins at the front) the film of air over the fins as the air goes through becomes heated and insulates the fins. Louvres introduce new leading edges - tourbulance the the pips and breaks up the insulating air and introduces fresh, cool air to the fins. I know that's how it works for passive airflow - could be that low speed fans are enough to introduce turbulance and break up the insulating layer anyhow. Not trying to preach to the converted btw - hoe it doesn't come across that way (I know you've done quite a bit of work on heatsinks)
 
i'd say no quieter how could it be any quieter than say 4 total quiet fans on air cooling ? the difference is extreme overclocks and keeping the silence
 
Well Im doing 3.5 ghz q6600 and getting acceptable temps with just the stock fan in the middle at 1000 rpm under a tuniq tower 120.

Extreme high end overclocks i agree with
 
Yep, on my old q6600 3.6ghz at 1.4 vcore, i was able to get away with quiet arctic cooling PWM fans, once i pushed it to 3.8ghz i had to invest in a couple of 2000 rpm silent eagles, and lap the cpu and my TRUE black, incidentally the temps at load were the same after theese changes, 68-62-68-62, my q9550 was a different matter, 3.8ghz on 1.216 vcore, very low speed sharkoon silent eagle se fans, (very rare, expensive and only 400-1200 rpm) quiet as a mouse and temps only hit 56-47-52-47, though i think the lowest cores on that cpu had stuck sensors.
 
Monkey Puzzle said:
I've heard the idea of air being saturated before - but if that were the case surelyy the air exiting the radiator would be at the samp temp as the water exiting the rad, and it isn't iirc.
Click to expand...

You are correct, and obtaining 100% air saturation would be near as damnit impossible (since as the temp delta between the two decreases the rate of heat transfer does as well). The issue isn't as such that the air is fully saturated after passing through a single layer of radiator, more that while the air may be only 60-70% saturated after a single layer of radiator getting higher levels of saturation and performance will take (unrealistically) large amounts more air-radiator contact time since the air-radiator temp delta is already fairly slim by this point.

Monkey Puzzle said:
The pa series radiators and similar deep widefinned ones don't perform above narrow-finned ones because there's a comparable or less surface area.
Click to expand...

Well the PA series do perform slightly better than single thickness radiators at low fan speeds, but that is due to the lower fin density of the PA rads to the single thickness ones (iirc all the single thickness rads are essentially all produced in the same factory as the swiftech MCR's, which have high fin densities) which allows greater airflow with low speed fans.

Of course with high speed fans high fin density radiators perform better since overall surface area plays a greater role. This is because with higher speed fans the air coming out of lower finned radiators isn't close to saturation and the larger surface area of the higher fin density radiators effectively allows the air to be saturated to a higher percentage.

Monkey Puzzle said:
The reason why i thought deeper radiators didn't do so well was that after the leading edge of the fins (where air first meets the fins at the front) the film of air over the fins as the air goes through becomes heated and insulates the fins. Louvres introduce new leading edges - tourbulance the the pips and breaks up the insulating air and introduces fresh, cool air to the fins. I know that's how it works for passive airflow - could be that low speed fans are enough to introduce turbulance and break up the insulating layer anyhow.
Click to expand...

While increasing turbulence in the radiators will help increase the saturation of the air, it will also decrease the actual 'flow rate' of the air through the radiator, so while you could quite a bit better saturation cooling performance may not actually increase. In a way the high fin density radiators would do much the same job as the various methods of increasing turbulence that you have suggested, and as we know different fin densities (and hence levels of turbulence) work best for different fan speeds.

Monkey Puzzle said:
Not trying to preach to the converted btw - hoe it doesn't come across that way (I know you've done quite a bit of work on heatsinks)
Click to expand...

haha, I've not really done any work with heatsinks, heck I've never run any experiments etc. I'm basing this mostly from theory and some from my experience with my own water loop. So I could be entirely wrong!
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom