Radiator stacking

[Moress]

So, if you have the extra room in the front of a case, is it beneficial to layer rads and fans? Like 2 360mm rads with fans in between them as well as on the front? Or will air temps make only the first rad work well?

[Beeeyeee]

It's not going to benefit you unless you have like 2 titanx'ps and a 6900k pumping heat into that thing. You'll be fine. Not to mention I don't think you can mount that properly

[Energycore]

I've wondered about this too. Just like you have two fin towers on big air coolers.

1 minute ago, Beeeyeee said:

It's not going to benefit you unless you have like 2 titanx'ps and a 6900k pumping heat into that thing. You'll be fine. Not to mention I don't think you can mount that properly

For very small cases though, where you can only do 1 120mm, maybe you can stack 120mms rads for the bit of additional cooling you need. It does require a custom loop however.

[LabRat]

short of using duct tape to hold the center together........... because you can't screw stuff together that way........ sounds interesting..

[LabRat]

1 minute ago, huilun02 said:

If a single loop goes through two rads, you're better off not stacking them. The bulk of the heat that is shed by the first rad could re-enter the coolant at the second rad... 

the whole loop will be the same temps anyways.

[fjdhdhcyfhf]

how does a boat with two outboard motors handle with two motors inline or two motors side by side?

[manikyath]

39 minutes ago, SCHISCHKA said:

how does a boat with two outboard motors handle with two motors inline or two motors side by side?

that isnt the right comparison, that'd be strictly talking about the airflow generated by stacking fans without other obstructions.

--

on topic: in theory, providing you have the pressure to push air trough the stack, and that the temperature delta between ambient and the liquid is high enough (take note, for the second rad in the stack, "ambient" would be the exhaust temp of the first rad) there is no reason why this wouldnt work.

 

that said, as noted already, there's only rare occasions where this would be necessary, as well as the temperature delta lowering also resulting in a lowering efficiency as you stack rads. on top of that, as someone already mentioned, the mounting difficulties.

 

for those who watched personal rig updates 2016-2170 (yes, thats a joke ) linus did, you can see he went for a "stacked rads" design in the front, but each rad has its set of fans, to overcome the static pressure requirement, and the loss in efficiency is something you just deal with in rackmount, because you cant really have an intake anywhere else than the front.

 

also, for those that have both a cpu and gpu aio cooler, if you mount one as intake, and the other as exhaust, you are essentially creating a "stacked radiator" design, just with the rest of your system in between.

[Energycore]

1 hour ago, LabRat said:

the whole loop will be the same temps anyways.

I feel like this is inaccurate. If the temperature of the whole loop was the same at all points, you wouldn't be carrying any energy away from the CPU

 

I think it's clear that the liquid at any point is hottest around the CPU block or similar, and coolest in one area of one of the radiators.

[manikyath]

6 minutes ago, Energycore said:

I feel like this is inaccurate. If the temperature of the whole loop was the same at all points, you wouldn't be carrying any energy away from the CPU

 

I think it's clear that the liquid at any point is hottest around the CPU block or similar, and coolest in one area of one of the radiators.

in the case of pc watercooling, the temperature delta trough the loop is very low, because the liquid can store a great amount of heat compared to what the chips kick out.

 

it would actually be really interesting for someone (jayztwocents? ) to make a custom loop with thermal sensors strapped to tubes in strategic locations (before and after the block, the rad, and the pump) and see what these numbers are *actually* like.

[done12many2]

25 minutes ago, Energycore said:

I feel like this is inaccurate. If the temperature of the whole loop was the same at all points, you wouldn't be carrying any energy away from the CPU

 

I think it's clear that the liquid at any point is hottest around the CPU block or similar, and coolest in one area of one of the radiators.

 

I monitor temps at two different places in my loop.  Right after the GPUs and another after exiting my lower radiators.  I can tell you that the absolute most I've ever seen is a 1c delta and that could have been within a margin of error for the sensors themselves.  

 

Water loops just don't have higher temps at any point in the loop as they equalize almost instantly.

[PCMasterDebater]

If you stack the Radiators you will have more overall cooling capacity from additional surface area. Heat transfer from the first radiator to the second wont make a lick of difference as the overall cooling capacity of the system has been essentially doubled.

 

Don't over think this. Cars have radiators on top of radiators on top of radiators and their cooling capacity is unaffected due to the overall capacity of the systems, regardless of the stacking.

 

hope this helps.

[corrado33]

4 minutes ago, PCMasterDebater said:

If you stack the Radiators you will have more overall cooling capacity from additional surface area. Heat transfer from the first radiator to the second wont make a lick of difference as the overall cooling capacity of the system has been essentially doubled.

 

Don't over think this. Cars have radiators on top of radiators on top of radiators and their cooling capacity is unaffected due to the overall capacity of the systems, regardless of the stacking.

 

hope this helps.

Completely incorrect. All of the radiators stacked on cars to go different systems. There is AC condenser (which looks like a rad.) then there is the radiator, then sometimes there is the oil cooler. All three of those go to completely different systems. 

 

You have not doubled the cooling capacity, and something like this would almost never be necessary. 

[PCMasterDebater]

1 minute ago, corrado33 said:

Completely incorrect. All of the radiators stacked on cars to go different systems. There is AC condenser (which looks like a rad.) then there is the radiator, then sometimes there is the oil cooler. All three of those go to completely different systems. 

 

You have not doubled the cooling capacity, and something like this would almost never be necessary. 

 

I'm not wrong. Since you now so well, can you explain to me how I have my engine coolant radiator, two transmission radiators, oil cooler, and as well as A/C condenser stacked on my drag car?

 

The two tranny coolers are stacked on top of the other, did it increase my transmission temperature?  

 

 

Nope, not in the slightest.

[KurtB]

18 minutes ago, PCMasterDebater said:

 

I'm not wrong. Since you now so well, can you explain to me how I have my engine coolant radiator, two transmission radiators, oil cooler, and as well as A/C condenser stacked on my drag car?

 

The two tranny coolers are stacked on top of the other, did it increase my transmission temperature?  

 

 

Nope, not in the slightest.

Not to question the results you had with your tranny coolers, but I'm sure the basic principles of thermodynamics would suggest that the air coming from the first radiator would have an affect on the second radiator. Much the same as the ambient air temperature has an affect.

[ToastyBear]

It is unresonable to say the second doesnt do anything. But reduced performance for sure.

 

Stacking 2 rads like that is effectively the same as running a single double thick rad with higher fan speeds. 

 

From benchmarks we know that double thick rads are only slightly better than single thick slim rads despite being up to 100% thicker. Triple thick monsta rads are just useless being no better than double thicks rads. Since doublr thick rads dont cost twice the price, its always worth it to buy them.

 

Just dont stack them. It barely does anything. 

[amvoith]

Here's a good read on the subject:

http://www.xtremesystems.org/forums/showthread.php?220874-More-Radiator-Sandwich-testing

 

TL;DR - no it doesn't increase cooling performance.

[Vigilante505]

This thread caught my eye because I have a spare XE360 Rad sitting around I don't know what to with. Was trying to think  of something useful for it instead of just sitting in the box. Now I know this isn't the way to go. 

[Qwweb]

The answer to why the system stays at nearly the same temperature is quite simple if you look at each radiator as a separate thermal system. Warm(er) water goes into the system and is cooled by the air (which takes away the heat) that air goes into the second radiator and re warms the water, but not to the same degree, as the water will reach equilibrium with the temperature of the air, at which point, heat transfer will no longer occur. The only time a decrease in temperature would occur would be if there was too much heat stored in the water to dissipate through the first radiator and at that point you would have much bigger problems.

[Qwweb]

On 1/26/2017 at 1:38 AM, manikyath said:

also, for those that have both a cpu and gpu aio cooler, if you mount one as intake, and the other as exhaust, you are essentially creating a "stacked radiator" design, just with the rest of your system in between.

Not really, as having space in-between the intake and exhaust allows for the air to (start to) go back down to room temperature helping a same size exhausting radiator deliver near-identical performance as an intake radiator of the same size with identical conditions (ie fans). Remember, solid/liquid to air heat exchange is inefficient, but same-phase exchange is much more efficient.

[manikyath]

1 minute ago, Qwweb said:

allows for the air to (start to) go back down to room temperature

if those two are the only intake and exhaust, heat doesnt magically disappear, it would only get hotter because of the rest of the system.

[Qwweb]

1 minute ago, manikyath said:

if those two are the only intake and exhaust, heat doesnt magically disappear, it would only get hotter because of the rest of the system.

Only in an isolated system (as stated in the first law of thermodynamics). Heat is always lost to entropy as any form of matter in the system will tend to absorb excess heat energy thereby contributing to the overall cooling by a small but noticeable margin (second law of thermodynamics). Perfect conservation of entropy only happens at absolute zero (third law of thermodynamics). So you are both right and wrong, as heat indeed does not disappear magically, but you are wrong in the fact that this is not an isolated system, so energy is not conserved; see first law. The continual heating that you mention implies a constant variable for entropy (no energy lost to the environment) which can only occur at absolute zero (no energy transfer can occur, different form equalibrium which is maximum entropy) thus breaking the second and third laws. 

Quote

The first law, also known as Law of Conservation of Energy, states that energy cannot be created or destroyed in an isolated system.

The second law of thermodynamics states that the entropy of any isolated system always increases.

The third law of thermodynamics states that the entropy of a system approaches a constant value as the temperature approaches absolute zero.

.

[manikyath]

3 minutes ago, Qwweb said:

any form of matter in the system will tend to absorb excess heat energy

except most surfaces in a system that are made to conduct heat are heatsinks significantly hotter than the air.

 

also, not isolated may be a bit of a stretch, seeing the popularity of sound dampening in cases, acryl not being a very good heat conductor, and a similar fashion for glass.

 

what you're talking about is theory, which is correct, but in practisce just a motherboard (excluding the cpu heat) is already more than what a case can evaporate without some form of airflow (active or passive)

[tempertantrum]

Radiator stacking is totally a thing. It's called countercurrent heat exchange, and depending on how you do it, you can get very different results.

https://www.quora.com/Why-is-a-counter-flow-heat-exchanger-better-than-a-parallel-flow-heat-exchanger
 

The key is 1.) to have the coolest air going over the coolest radiator, and 2.) to get enough air flow over the radiators so that the air hitting the second one still has enough of a heat differential to cool it. The only comprehensive experiment of this I've found used crappy fans (Nanoxia FX-1250) for the task. I actually just asked Jayztwocents to do this experiment with better equipment, especially fans. So... let's hope and see?