Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
KingMongkut

Why radiator on the Intake and not Exhaust?

Recommended Posts

Posted · Original PosterOP

Hello!

I just always wondered why the radiators for cpu coolers are placed on the intake to the case; doesn't that just send the hot air into the PC again? Why not on the exhaust?

Thanks so much for your time and wisdom!

Link to post
Share on other sites

with the radiator as intake, the CPU will always get fresh air, while the GPU receives the heated air. With the radiator as exhaust, the CPU will receive the hot air off the GPU.

 

Sometimes, one works better than the other, it varies greatly based on case and airflow.


I WILL find your ITX build thread, and I WILL recommend the SIlverstone Sugo SG13B

 

Primary PC:

i7 8086k (won) - EVGA Z370 Classified K - G.Kill Trident Z RGB - Force MP500 - Jedi Order Titan Xp - The venerated Hyper 212 Evo (with RGB Riing flair) - EVGA G2 650W - Black and green theme, Razer branwashed me.

Draws 400 watts under max load, for reference.

 

Linux Proliant ML150 G6:

Dual Xeon X5560 - 24GB ECC DDR3 - GTX 750 TI - old Seagate 1.5TB HDD - Dark moded Ubuntu (and Win7, cuz why not)

 

How many watts do I need? Seasonic Focus threadUserbenchmark (Et al.) is trash explained, PSU misconceptions, protections explainedgroup reg is bad

Link to post
Share on other sites
6 minutes ago, KingMongkut said:

I just always wondered why the radiators for cpu coolers are placed on the intake to the case; doesn't that just send the hot air into the PC again? Why not on the exhaust?

I believe its so the cpu is cooled with the cooler air from outside the case instead of the warm air thats leaving the case which wouldnt work as well.


I am far from an expert in this so please correct me if I’m wrong.

Quote or tag me so I can see your response

 

PSU Tier List

Motherboard Tier List

PARROT GANG

Mentioned in 7/10/20 WAN Show

Mentioned in 7/15/20 Techlinked

Mentioned in 7/17/20 Techlinked

Mentioned in 7/31/20 WAN Show

Mentioned in 7/31/20 Techlinked

Link to post
Share on other sites
6 hours ago, KingMongkut said:

But isnt the heat from the cpu taken by the liquid? Does the air outside the heat sink matter?

The heat from the cpu is taken by the liquid, but the liquid itself needs to be cooled.


I am far from an expert in this so please correct me if I’m wrong.

Quote or tag me so I can see your response

 

PSU Tier List

Motherboard Tier List

PARROT GANG

Mentioned in 7/10/20 WAN Show

Mentioned in 7/15/20 Techlinked

Mentioned in 7/17/20 Techlinked

Mentioned in 7/31/20 WAN Show

Mentioned in 7/31/20 Techlinked

Link to post
Share on other sites

They aren’t always like that. That’s a simple choice just like everything else. 
Id never do that myself unless I was desperate for space and I would change it later. 


Main RIg Corsair Air 540, I7 8700k, ASUS ROG Strix Z370-H, G.Skill TridentZ 16GB, EVGA 1080TI SC Black, EVGA 850 GQ, Acer xG270HU 2560x1440@144hz

 

Spare RIg Corsair Air 540, I7 4770K, Asus Maximus VI Extreme, G.Skill Ares 32Gb, EVGA 1080sc & 1060 SSC, Corsair CX850M, Acer KG251Q 1920x1080@240hz 

 

Link to post
Share on other sites
10 hours ago, KingMongkut said:

Hello!

I just always wondered why the radiators for cpu coolers are placed on the intake to the case; doesn't that just send the hot air into the PC again? Why not on the exhaust?

Thanks so much for your time and wisdom!

Basically unless you have a very powerful CPU, the primary source of heat in your system will be your GPU. Even when your CPU is under load, the temperature of the liquid itself in the AIO doesnt get that hot and therefore the air going through the radiator isn't considerably hotter either.

 

So if you have the rad as the intake the temp of air going is still only a couple degrees over ambient at most so the cooling of other components isn't affected as much. On the other hand, if you keep is as exhaust, your GPU will heat up the air inside the case a considerable amount degrading the CPU cooling.

 

See this video for more details.

Link to post
Share on other sites

It really depends purely on the airflow in your case to begin with, and of course where the radiator is mounted.

 

A front mounted radiator (like if that's the only place you can fit it) is always going to be intake.

 

With a top mounted radiator: 

If you have excellent airflow from the front of the case with powerful fans (e.g. a mesh type or unrestricted or open front panel or open door), then the top rad being exhaust will help the hot air leave the case and the warm air from the radiator vent out, since hot air rises.

 

If you have a 'non airflow' case, like some of the extremely large cases, with a partially blocked front intake panel (like a non-mesh door that is closed), where the fans air intake is limited, then top mounted radiator as intake will help compensate for the front impeded airflow and can lower temps on both your CPU and RAM quite a bit.  This type of setup works best with high CFM/static pressure radiator fans (like those powerful Noctua industrial fans).

 

A fractal design 7 XL with the front door closed works best with front and top fans as intake, since the front door restricts airflow and the vents on the side prevent high CFM fans from pushing air effectively, unless the front door is opened or removed (which turns it into a large mesh type/serrated case)

Link to post
Share on other sites

It all depends on your case, the fan layout and your components.

 

The cooler the air that goes through the rad the greater the heat difference ΔT between liquid temperature and air temperature becomes. The bigger the ΔT the more efficient the heat transfer becomes. Because that's all you do: you transfer heat from the CPU's IHS to the water through TIM and coldplate and then from the Liquid to the ambient air through the radiator's surface.

 

Edit: scenarios below assume radiators as exhaust since those are being influenced by your case temps. When radiators are placed as intake there's no difference (ignoring the small amounts of heat radiation).

 

If you have only your CPU on water and the GPU is on air heating up the case then you will reduce your ΔT making heat transfer less efficient. If the hot air of the GPU is being pulled out efficiently (or if your case has a GPU compartment connected with a riser card) the effect will be marginal.

 

If you have both GPU and CPU on water, the effect will be small since you will only have your VRMs heating up the system which should not be too hard to cope with with a few solid case fans on low revs. 

 

If you have GPU, CPU and VRMs on water I'd say it really doesn't matter anymore because there's hardly any bigger heatsource that is not cooled anymore.

 

Next thing is restrictiveness of the fan/radiator mounts. There's no point in putting your radiator at a place where the fans need to work overtime to get some tiny bit of airflow through the radiator. A very restrictive fan mount already cries for static pressure fans (or very high revs). When adding a radiator on top of that and you'll end up with very poor airflow and thus little heat transfer because your radiator needs air going passed its fins in order to transfer heat. Less air going through = rising temperature of that air = smaller ΔT = less efficient heat transfer + higher case temps.

 

Bottom line: there is no general rule where to put it, it depends on the case you're building and you will need to test configurations out in order to find the best case. 

 

P.S.: hot air rising is true in the open when there's a heat gradient, it's a negligible to non existent effect in most pc cases though because the gradient is negligible. Common pc cases are just too small for this to work and usually have too little of a heat gradient. The reason for warm air rising is its lower density. A ΔT of 5K (25°C vs 30°C) changes the density by only ~1.65%. You think you have more than a 5K gradient in your system? The airflow created by your fans is far bigger than any turbulences caused by a tiny heat and density gradient.


Use the quote function when answering! Mark people directly if you want an answer from them!

Link to post
Share on other sites

If you have a side vent the GPU will pull fresh air from there.  Hot air rises, so the top should always be the exhaust.  Cold air is more dense so it settles to the bottom.  

As a professional pilot we deal with this sort of thing all the time, a 5c degree gradient change is actually about 3,000 ft of altitude so there is a large change in vertical motion.  The cool air coming in the PC will disrupt the warm air so there are channels and vortices within the case.  As the cool air enters it will fall in the case and cause a cyclonic rotation of the warm air mass (counter clockwise)  so that air system naturally rises and sucks in higher pressure from the bottom while expanding vertically.  Just because we are talking about large air masses does not mean the same physics apply in small spaces.  Everything follows the same physics just on different scales.  It's the same with the vortex on the edge of a wing, about the size of a computer case on small wings. 

Thus, the exhaust fan should always be on top to rid the hottest air, and to aid in the natural movement and flow of the air. 

Link to post
Share on other sites
2 hours ago, daveholland86 said:

If you have a side vent the GPU will pull fresh air from there.  Hot air rises, so the top should always be the exhaust.  Cold air is more dense so it settles to the bottom.  

As a professional pilot we deal with this sort of thing all the time, a 5c degree gradient change is actually about 3,000 ft of altitude so there is a large change in vertical motion.  The cool air coming in the PC will disrupt the warm air so there are channels and vortices within the case.  As the cool air enters it will fall in the case and cause a cyclonic rotation of the warm air mass (counter clockwise)  so that air system naturally rises and sucks in higher pressure from the bottom while expanding vertically.  Just because we are talking about large air masses does not mean the same physics apply in small spaces.  Everything follows the same physics just on different scales.  It's the same with the vortex on the edge of a wing, about the size of a computer case on small wings. 

Thus, the exhaust fan should always be on top to rid the hottest air, and to aid in the natural movement and flow of the air. 

This is logical, but if the intake in the front isn't strong enough, then top intake is needed.

I switched from top exhaust to top intake on a top mounted radiator, and dropped 4C temps on the CPU.  My video card is a reference blower so that helps exhaust that air out, otherwise I may have slightly worse results.

Link to post
Share on other sites
4 hours ago, daveholland86 said:

If you have a side vent the GPU will pull fresh air from there.  Hot air rises, so the top should always be the exhaust.  Cold air is more dense so it settles to the bottom.  

As a professional pilot we deal with this sort of thing all the time, a 5c degree gradient change is actually about 3,000 ft of altitude so there is a large change in vertical motion.  The cool air coming in the PC will disrupt the warm air so there are channels and vortices within the case.  As the cool air enters it will fall in the case and cause a cyclonic rotation of the warm air mass (counter clockwise)  so that air system naturally rises and sucks in higher pressure from the bottom while expanding vertically.  Just because we are talking about large air masses does not mean the same physics apply in small spaces.  Everything follows the same physics just on different scales.  It's the same with the vortex on the edge of a wing, about the size of a computer case on small wings. 

Thus, the exhaust fan should always be on top to rid the hottest air, and to aid in the natural movement and flow of the air. 

While I agree with most of your post, I cannot agree with your conclusion. You mention the vortices yourself. The size of the system does have an impact - especially in regards to mixing of air volumes of different temperatures. You can make use of a thermal column (or a sort of chimney effect) if the case is designed accordingly. As soon as you're blowing air in and out from the sides the kinetic energy of the air will outperform the turbulent air you have in a regular case and there won't be a natural flow from bottom to top due to a thermal column anymore. There would be rising air if you wouldn't disturb the air with your airflow from the fans - but that's usually the case.

 

I specifically gave the numbers for air with a 5K gradient - I would not expect such a big gradient in a common case (and from my information, a 5K gradient would equate to ~2500ft but that's just me being nitpicky ;) and of course it depends on the specific conditions; in a strong thermal column it might even be just 1200ft). 

 

Note: many many approximations below!

 

How much of a gradient can we expect? This depends on the specific heat sources in a system. Assuming a case of ~40l (so .04m³) and 3x 120mm intake fans (randomly choosing beQuiet Silent Wings 3 PWM fans because I like them) then at max speed these 3 fans could theoretically move ~4m³ of air per minute or 100x the volumne of that case. Of course the restrictiveness of the case would reduce this (only intake creates a slight positive pressure inside the case) but then exhaust fans would increase this again. I'm not a physicist nor am I an engineer so I can't give any figures now how two fans would interact with one intake and one exhaust in an enclosed case - no idea at this point wether they'd add up to some point or wether they just operate at their maximum. I'd expect some interaction but definitely not in a linear fashion. "But I digress!" Let's assume the real world scenario with an effective airflow of ~1m³/min. That's still 25x the case volume. That would be 100% exchange every 2.4s.

 

Now, how much heat do we have in that case? Ignoring CPU and GPU for now one of the main sources are the VRMs (~20-30W peak?), your NVMe (about 5W-ish?) and that's mostly it since the PSU in most cases is in a more or less separated compartment and often blows it's hot air out directly not interfering with the rest of the case. So maybe 35W of heat power at absolute peak load. With those heat sources it would take ~1.5s (ish) to heat up the air volume inside of a 40l case by ~1K - assuming perfect heat transfer from the heatsinks - which is not the case as heat transfer depends on the temperature gradient the surface of the heatsink and to some extent airflow. 

 

Considering the airflow through the case the gradient might be at max ~1.5-2K considering max heat output during a torture test and perfect heat transfer. We haven't yet thought about how the air volumes mix due to vortices and turbulences.

 

Sure, your GPU would potentially add more heat to the equation but the amount is heavily dependant of the cooler design and the layout of the specific case. 

 

We could now try to calculate the speed of the rising air by some approximations and by considering the moment of intertia of that air volume we could estimate how much airflow would be impeded - assuming a perfect updraft without vortices. If we'd want to consider those I guess we'd be better off by using some CFD module in Fusion 360 and a 3D model of a specific PC case. If I had to guesstimate, I'd say low single digit percentage.

 

Bottom line: your airflow is many times more powerful than any updraft generated by a thermal column as a result of a temperature gradient.


Use the quote function when answering! Mark people directly if you want an answer from them!

Link to post
Share on other sites
8 hours ago, daveholland86 said:

If you have a side vent the GPU will pull fresh air from there.  Hot air rises, so the top should always be the exhaust.  Cold air is more dense so it settles to the bottom.  
 

 

True scientific fact yes, but when it comes to multiple high power fans doing work in a confined space, that whole aspect of it is negligible.

Link to post
Share on other sites
3 hours ago, bowrilla said:

Bottom line: your airflow is many times more powerful than any updraft generated by a thermal column as a result of a temperature gradient.

It depends on the size of the thermal column, we have some that move an 85,000 lb aircraft.  I am just saying if we position our fans to aid in the natural movement it would be most efficient.  You can use fans and exhaust through the bottom but you are working against natural tendencies and thus will require more energy and you will have leaks and hot air circling back around.  The idea is to have the most powerful fans pushing air in while the exhaust fan does not need to be as powerful because there is a natural flow already moving that way.  I also cant feel air movement on my hand 8 inches behind a radiator with a 50 CFM fan on it.  Air is resistant to change, thus how we make thrust with propellers the act of deflecting air rearwards pulls the aircraft forwards through the air. 

This of course assumes it's a close case.  If the case is open then the cold air mass will want to fall on it and the not air pushed out the sides of the bottom.  

 

Of course, we are a little too far in theory of it.  You simply put fans on it and it works in practice. 

Link to post
Share on other sites

Its a tiny box. Doesnt really matter overall. Just comes down to priorities. Like if you did it the right way but having it on the top as exhaust and use a blower card. Both get fresh air and wont compromise each other. Or use an open air cooler if thats your thing. Thus putting the gpu above the cpu as you should. Gpu lowers its clocks way before a cpu ever will. 


Main RIg Corsair Air 540, I7 8700k, ASUS ROG Strix Z370-H, G.Skill TridentZ 16GB, EVGA 1080TI SC Black, EVGA 850 GQ, Acer xG270HU 2560x1440@144hz

 

Spare RIg Corsair Air 540, I7 4770K, Asus Maximus VI Extreme, G.Skill Ares 32Gb, EVGA 1080sc & 1060 SSC, Corsair CX850M, Acer KG251Q 1920x1080@240hz 

 

Link to post
Share on other sites
On 7/14/2020 at 12:06 AM, KingMongkut said:

I just always wondered why the radiators for cpu coolers are placed on the intake to the case; doesn't that just send the hot air into the PC again? Why not on the exhaust?

I've always placed mine as exhausts.

Link to post
Share on other sites
10 hours ago, daveholland86 said:

It depends on the size of the thermal column, we have some that move an 85,000 lb aircraft.  I am just saying if we position our fans to aid in the natural movement it would be most efficient.

Exactly, the thermal column is very tiny incomparison with just ~.04m³ in comparison to several thousand m³ of air moving (a cube of 10m edge length has a volume of 1,000m³ - that's still tiny in comparison of thermal columns moving airplanes). And you are correct, it would be most efficient working with a theoretical thermal column inside the case, the effects will be absolutely negligible though and most likely within margin of error.

 

10 hours ago, daveholland86 said:

Air is resistant to change, thus how we make thrust with propellers the act of deflecting air rearwards pulls the aircraft forwards through the air. 

Yep, that's Newton's 3rd Law. You're moving/accelerating a certain volume and therefore mass of air in one direction (actio) therefore you experience an equally large force into the opposite direction (reactio).

 

10 hours ago, daveholland86 said:

This of course assumes it's a close case

We also assume it's a perfect box with a perfectly insulated case, no further disturbances of airflow or blockages, etc. 

 

10 hours ago, daveholland86 said:

Of course, we are a little too far in theory of it.  You simply put fans on it and it works in practice. 

Very much indeed. It's a fascinating topic to think about and maybe even experiment with, in practice the effects are small to negligible. You slap fans on it and it works unless you placed them in a very illogical and unintuitive way. Exhaust at the top often works very well - mainly because top vents are often very unrestrictive and placing a radiator there often gives you the space for the biggest rads. The temperature difference between exhaust and intake will most of the time not be that great to the point of your loop suddenly not working anymore. This was just fun to think about.


Use the quote function when answering! Mark people directly if you want an answer from them!

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×