Understanding case airflow with top mounted intake radiator

[harisund]

I have Googled a lot about this. Every post I see has a standard response, "it is ok as long as your case has good exhaust".

What the heck does good exhaust mean?

Here are my thoughts on the subject. I have seen pictures or air flow where cold (outside) air is brought in typically from the front and side ... becomes hot as it passes over the components ... hot air rises .. exhausts out the back and the top. All makes sense so far.

When you mount a dual 120/140mm radiator on top as intake, you are forcing cool air to blow downwards, heating in the process. Additionally, you are bring the hot air now over the other components in the motherboard, which I would think is far worse than a stock fan since a stock downwards facing fan actually also helps in cooling the MB components.

If I am understanding right, basically you are intaking from just about everywhere there's a fan, and forcing all the exhaust through whatever vents you have in the system, and one rear exhaust fan (which I have never seen to be bigger than 140mm).

So what's the secret? How does this make it a good idea? I can understand in something like a carbide air 540, yea you have a "wall of air" being pushed, so it might be fine, but in your normal cases, aren't you actually going against convection .. and logic?

 

[aigomorla]

the delta from rad in to rad out will not variate more then a couple of C's unless you got a super efficient radiator which blows all conventional radiators out of the water, or you have your fans so low that the air is getting super saturated.

Or you have way too much heat source and insufficient radiator capacity which causes your coolant temp to be extremely high.

So rule of thumb... you always want the freshest air source to enter the rad, and the air out wont be that much hotter then the air in.

 

[Tweakin]

...So rule of thumb... you always want the freshest air source to enter the rad, and the air out wont be that much hotter then the air in.

Agreed!

As for the op, I've seen lots of folks talk about "positive" case pressure vs "negative" case pressure as it relates to thermals and dust buildup. I believe that the case itself "leaks" so much air that as long as you have movement all is well. Add to this the type of fans you are using, and you could end up with a case that "stagnates" due to lack of air flow. I believe this is more of an issue low cfm or silent configurations as most "air" cooled systems move adequate air through the case to keep the gpu's cooled. I also believe that front-low intake and top-high exhaust is the most efficient as it utilizes convection, however minimal.

In my current build I'm going to intake via my front rad (140x3) and exhaust through the top rad (120x2) in an effort to create better air flow in the case. I will also reverse the top rad to intake to see what the delta is, if any.

 

[BonzaiDuck]

Agreed!

As for the op, I've seen lots of folks talk about "positive" case pressure vs "negative" case pressure as it relates to thermals and dust buildup. I believe that the case itself "leaks" so much air that as long as you have movement all is well. Add to this the type of fans you are using, and you could end up with a case that "stagnates" due to lack of air flow. I believe this is more of an issue low cfm or silent configurations as most "air" cooled systems move adequate air through the case to keep the gpu's cooled. I also believe that front-low intake and top-high exhaust is the most efficient as it utilizes convection, however minimal.

In my current build I'm going to intake via my front rad (140x3) and exhaust through the top rad (120x2) in an effort to create better air flow in the case. I will also reverse the top rad to intake to see what the delta is, if any.

Aigo knows I'm an H2O bystander.

With air-cooling, pressurizing the case can only help -- provided that the air channels quickly to the parts that need cooling and then swiftly exhausted.

You'd think with "a couple degrees" difference between rad intake and exhaust, it wouldn't much matter having a radiator placed coincident with exhaust. But intake would be more ideal.

Without a lot of detail -- and someone inquired for me to produce pictures which I'll do eventually -- I'd built a huge case back in '07 with dual-140 bottom intake fans and plenty of room for a radiator with dual front-panel 120's. That works -- if you put your case on 3" double-braked-caster wheels as I did.

As for top-mounted radiators, even as a contemplative bystander I keep having dreams of "rain" inside my case. Just a thought . . . if you can call it that . . .

 

[Tweakin]

...As for top-mounted radiators, even as a contemplative bystander I keep having dreams of "rain" inside my case. Just a thought . . . if you can call it that . . .

Funny you say that as I have always had dreams of 600gram heat sinks flexing the board to much...and that's why I'm all washed up! (note attempt at humor).

Sorry to the op for stealing the thread...

 

[james1701]

What about something like a 900D and air cooled graphics cards, with a large rad on top.

I just found a great deal on this case, but my 7970ghz cards are nonreference. I want to wait until I can put a full block on them.

I am going to go ahead and setup all the rads for cpu and gpu, so I don't need to do it later. Im getting a 480 on top, and a 480 on the bottom.

With no side fan like the HAF, would intake be best on both rad's to keep the case flushed for air cooled cards?

 

[BonzaiDuck]

Funny you say that as I have always had dreams of 600gram heat sinks flexing the board to much...and that's why I'm all washed up! (note attempt at humor).

Sorry to the op for stealing the thread...

I think at one time the Intel spec on their boards was 400gm -- probably overly cautious, given the products developed over the years, improvements in the boards themselves and so on. Using double fans can add 3/4 lb. to the weight -- and torque. But I never heard of anyone with a large air-cooler who damaged their mobo with it.

This year, I'm "testing the water" so to speak, but jumping into it is no certainty.

ADDENDUM: I'll add something to this, because I'm constantly dreaming up "new computer designs" even if I don't act on them. I'd like to minimize the fan-business with water-cooling. With casual searches, I'd found square rads -- maybe even "thick" ones -- that fit 200mm fans. That is -- one low rpm moderate-to-high output. But it only works with certain cases. Since I recycle my cases in my own tradition of using old 90s-era towers and modding them, I keep imagining ways to adopt water-cooling for my HAF midtowers. I don't want full tower cases anymore, but I used to prefer them.

 

[BonzaiDuck]

. . .

So what's the secret? How does this make it a good idea? I can understand in something like a carbide air 540, yea you have a "wall of air" being pushed, so it might be fine, but in your normal cases, aren't you actually going against convection .. and logic?

If you can't put waterblocks on everything -- and I'd probably just do it to the CPU and skip water-cooled graphics while others may actually need the latter -- I could still have an agenda for air-cooling all the other components, and there are ways to do that effectively according to the prevailing idea of case airflow.

And it seems pretty clunky to me to use the top case vent(s) for intake. With good intake from bottom, sides and front, I wouldn't see any problem putting a rad on a vent that exhausts air. My initial choice, though, would have the rads in the side, bottom or front where they intake air into the case.