Rad Thickness

[Dabasepc]

Hi,

 

Just wondering, for a 240mm Rad, thickness between 20, 38, 40+, what temps do you get vs each other? how much better is it?

[airdeano]

usually the same across the thickness as the fan/rad combo is the same. what

usually happens is the time it takes the system to stabilize in temperature increases.

assuming the fan speed is the same for all tests. only thing that lowers temperatures

in more airflow through the radiator and that is done with more radiators and fans.

but eventually, the rate of temp drop to more radiators gets to a rate of non-return.

 

20 fans/radiators might only cool .5° cooler than 10 fans/radiators. but it will take

longer for the system to reach a stable temperature.

[Enderman]

usually the same across the thickness as the fan/rad combo is the same. what

usually happens is the time it takes the system to stabilize in temperature increases.

assuming the fan speed is the same for all tests. only thing that lowers temperatures

in more airflow through the radiator and that is done with more radiators and fans.

but eventually, the rate of temp drop to more radiators gets to a rate of non-return.

 

20 fans/radiators might only cool .5° cooler than 10 fans/radiators. but it will take

longer for the system to reach a stable temperature.

That's not how it works. It's not only about the rate of airflow, but also about the surface area for heat exchange.

A thicker radiator will cool quite a lot better than a thin radiator. Also, a higher density fpi radiator will cool better than a low density fpi radiator.

Even with the same fans at the same rpm.

 

Look at the H105 vs H100i for example. The H105 is only a few millimeters thicker than the H100i and that already gives it 3-5C lower temperatures.

[airdeano]

..a higher density fpi radiator will cool better than a low density fpi radiator.

Even with the same fans at the same rpm.

 

mmm, the higher density vs lower density is not explained correctly.

 

low FPI radiators perform using less airflow (air rate/rpm) will cool better than

the high FPI using the same lower rpm. but the low FPI will start to lose some

cooling performance as the high FPI start to get better with more fan speed.

 

the alphacool is low FPI, Swiftech/HL GTX is high FPI as explained by

martinsliquidlab.org

 

top 11 in the 600rpm chart are sub 12 FPI. the BIX is 25% less dissipation than

the UT60 (both being 60mm thick)

 

note in the 600rpm the low FPI rate higher in watt dissipated over the higher

FPI radiators. only till you get to the 1400 rpm range you see the BIX getting

better in the performance. improvement comes to 1000 rpms to 17% spread,

at 1800rpm, the spread between the two gets to 8% difference. only thing

changed was the fan speed.

 

until 2200 rpm, the BIX is in the 33% bracket better overall (5%) from the

lower rpm models.

 

 

source: http://martinsliquidlab.org/2012/04/14/360-radiator-shootout-summary/

 

you'll also find similar result within the Bundymania shootout, too: http://www.techpowerup.com/forums/threads/bundymania-user-review-triple-radiator-360-roundup-with-22-rads.173028/

[jeffrey_rodriguez]

Hi,

 

Just wondering, for a 240mm Rad, thickness between 20, 38, 40+, what temps do you get vs each other? how much better is it?

hi dabasepc,, please forgive me geeting into your question,,but i would like to add some more..

 

usually the same across the thickness as the fan/rad combo is the same. what

usually happens is the time it takes the system to stabilize in temperature increases.

assuming the fan speed is the same for all tests. only thing that lowers temperatures

in more airflow through the radiator and that is done with more radiators and fans.

but eventually, the rate of temp drop to more radiators gets to a rate of non-return.

 

20 fans/radiators might only cool .5° cooler than 10 fans/radiators. but it will take

longer for the system to reach a stable temperature.

hi aiedeano,, you said that it would be the same across the thickness as the fan/rad combo,, i currently have a 360mm,60mm thick rad and a 240mm 30mm thick rad rad on my system.. would you say that if i remove the 360mm rad ,i would get the same temperature OR have a slightly higher temperature?

 

That's not how it works. It's not only about the rate of airflow, but also about the surface area for heat exchange.

A thicker radiator will cool quite a lot better than a thin radiator. Also, a higher density fpi radiator will cool better than a low density fpi radiator.

Even with the same fans at the same rpm.

 

Look at the H105 vs H100i for example. The H105 is only a few millimeters thicker than the H100i and that already gives it 3-5C lower temperatures.

hi enderman,, would you say that at xflow 240mm 20mil thick rad with high fpi will perform the same as the low fpi ek 360mm, 60mil thick rad? let say we would use the same fans @ the same rpm's.

[airdeano]

i currently have a 360mm,60mm thick rad and a 240mm 30mm thick rad rad on my system.. would you say that if i remove the 360mm rad ,i would get the same temperature OR have a slightly higher temperature?

 

dunna know your build specs, but removing the 360x 60 would deffo increase

your temperatures. i am assuming you have a CPU/GPU loop?

[Enderman]

mmm, the higher density vs lower density is not explained correctly.

 

low FPI radiators perform using less airflow (air rate/rpm) will cool better than

the high FPI using the same lower rpm. but the low FPI will start to lose some

cooling performance as the high FPI start to get better with more fan speed.

 

the alphacool is low FPI, Swiftech/HL GTX is high FPI as explained by

martinsliquidlab.org

 

top 11 in the 600rpm chart are sub 12 FPI. the BIX is 25% less dissipation than

the UT60 (both being 60mm thick)

 

note in the 600rpm the low FPI rate higher in watt dissipated over the higher

FPI radiators. only till you get to the 1400 rpm range you see the BIX getting

better in the performance. improvement comes to 1000 rpms to 17% spread,

at 1800rpm, the spread between the two gets to 8% difference. only thing

changed was the fan speed.

 

until 2200 rpm, the BIX is in the 33% bracket better overall (5%) from the

lower rpm models.

 

source: http://martinsliquidlab.org/2012/04/14/360-radiator-shootout-summary/

 

you'll also find similar result within the Bundymania shootout, too: http://www.techpowerup.com/forums/threads/bundymania-user-review-triple-radiator-360-roundup-with-22-rads.173028/

 

the performance results from the combination of two factors 1) surface area and 2) airflow is not a linear equation, it is quadratic

 

at a certain fan rpm, there is a point of maximum performance between high surface area (which causes less airflow) and low surface area (causing high air flow)

high surface area can be radiator thickness or fpi, they both scale linearly

 

the "balance" is the point where you get the lowest temperature by having a radiator density that doesnt restrict the airflow too much, yet enough surface area to dissipate the heat.

this point is basically impossible to determine without me having every single radiator thickness and fpi

 

but you can see that my theory is true due to the fact that with the same fans, at the same rpm, you will get the best performance from a moderate fin-density and moderate thickness.

a low fpi thin radiator will perform worse even though it gets lots of airflow

a high fpi extremely thick radiator will also get worse performance because of extremely restricted airflow

 

this tells us that it is a quadratic curve, similar to this and the curve changes for every rpm, cfm, and static pressure of every fan

[Enderman]

hi enderman,, would you say that at xflow 240mm 20mil thick rad with high fpi will perform the same as the low fpi ek 360mm, 60mil thick rad? let say we would use the same fans @ the same rpm's.

this depends on the fpi we are talking about

 

you can compare the performance by doing 240mmx120mmx20mmx30fpi = 17mil

vs 360x120x60x15fpi = 39 mil

 

so you can see that if the same amount of air was moving through the radiator, the second option would give approximately 2.5x the performance

now this isnt the case because the airflow will slow down going through the thick radiator, but you do have 150% airflow due to a third fan

 

in clonclusion, the triple radiator will be quieter while performing better (but it will not be 2.5x better, that is just theoretical peak)

[jeffrey_rodriguez]

dunna know your build specs, but removing the 360x 60 would deffo increase

your temperatures. i am assuming you have a CPU/GPU loop?

yes I am currently running a radeon 6990 and a i5 4460 on a single loop. i was planning to remove one of my rads ,but I am having a bit o a doubt and hesitation