Liquid cooling - should it pull the air INTO the case or push out?

Status
Not open for further replies.

Lord Northern

Prominent
May 26, 2017
7
0
510
Hi.
This isn't the classic push vs. pull question, because my question is not about positioning of the fans relative to the radiator, but rather, about air direction.
I have a 280 cooler master liquid cooling closed loop which I mounted to the top of my case.
Currently, the fans pull air from the case through the rad, and into the environment.
Is it correct, though? Wouldn't it be better if I rotated the fans so that they would pull air through the rad into the case?
I do have sufficient airflow in my case. I have two front fans pull in, and another one in the back pulling air right to the rad.

So, what should I do then?
 
Solution
This is a classic question:
Whether tis better to draw cool air in through a front mounted rad, or push air out a top mounted rad.

Both have advantages and disadvantages.

Front intake:
Cooler air coming through the rad, so, in theory, better cooling of the liquid that cools the CPU.
However, this blows prewarmed air into the case.

Top exhaust:
Blows directly out the top.
However, this is using air that has already been warmed a little by the other components. GPU, etc.

I posit that the individual case, cable management, and components is the key factor. There is no One True Way.

I have my Cryorig A80 blowing out the top.
I would run it as exhaust (air out) simply because it would heat up the rest of the case otherwise (such as the GPU and the VRM which has little cool airflow to begin with).

That being said, I would recommend you push the air out the radiator, not pull. Its not as efficient on the fans.
 

USAFRet

Titan
Moderator
This is a classic question:
Whether tis better to draw cool air in through a front mounted rad, or push air out a top mounted rad.

Both have advantages and disadvantages.

Front intake:
Cooler air coming through the rad, so, in theory, better cooling of the liquid that cools the CPU.
However, this blows prewarmed air into the case.

Top exhaust:
Blows directly out the top.
However, this is using air that has already been warmed a little by the other components. GPU, etc.

I posit that the individual case, cable management, and components is the key factor. There is no One True Way.

I have my Cryorig A80 blowing out the top.
 
Solution
Good analysis from USAFRet.

I might add that unless all of your case intake air comes from one source only(usually the front) and is filtered, other options will draw in unfiltered air from adjacent openings and you will accumulate dust on your parts.
 

svx94

Prominent
Sep 24, 2017
5
0
510


Allow me to "reactivate" this topic, since I have the same questions, and agreed with geofelt's answer. However, I see so many people doing the opposite (doesn't made them correct by default). so, I'd like to hear more opinions.

I can imagine it you are run out of space to mount radiators, and take both front and top with radiator, there is no intake any more. In that case, it might be better off to have the front as intake, even the air get warmer through the radiator - better something (airflow, even a little warm) than nothing? I think too many people go for the look rather than function, even the manufactures.

This question is driving my selection of case for my coming Xmas build, and the first one since college time, and thing are dramatically different now, with water cooling, etc. I plan to have EK PE360 loop to take care of CPU and GPU (with build-in WB). I learned I still need to cool VRMs, memory, M.2 etc. My current plan is to have the TT Core V71 tg, which has 2x 200mm intake in the front to supply cool air into the case. 140mm back exhaust, 360 rad on top also exhaust. I also plan to have the PSU fan facing up, so it suck air to increase the flow. The problem is scalability, in case I want to add another rad.
 

svx94

Prominent
Sep 24, 2017
5
0
510


Thanks for sharing! It is very interesting input.

I do have several doubt about his conclusion:
1. Assuming everything in the test is 100% accurate, he only considered CPU and GPU. In fact, if you water cool both of them, the temp. is more defined by your installation quality (TIM application, e.g.) and the capacity of the cooler. The real purpose of the airflow is to cool the REST, which is not measured at all;
2. I do question the conclusion. Not sure he reinstall the CPU block for different config or not; Because the TIM usage can result in huge difference;

The physic doesn't change. To have a good cooling, you need transfer the heat effectively with several factors:
1. Delta-T, the bigger difference of the temp, the better cooling. Blowing warm air (vs. cooler room temp. air) is against it.
2. Faster flow, if the air moves, the heated up air will not stay around, and result in better cooling;
3. Better TIM. Air is the worst but cheapest. Unless you have waterblock for everything, you still need airflow - which is the reason we are discussing here;

One factor I didn't think before is that when you use rad as intake, the cooling effectiveness of the loop is maximized, which can result in lower CPU temp. When you use rad as exhaust, the "intake" for the rad is from inside, which can be warmed up by the VRM, etc. So, you trade the CPU cooling with VRM cooling.
 

svx94

Prominent
Sep 24, 2017
5
0
510


Thanks! What you said make a lot of sense to me!

I think I will try to have strong intake to overwhelm level, so the rad take less warmed air due to VRM, etc.
 
Status
Not open for further replies.