cooling

[pete g]

hi, i want to cool an enclosure. is it better to blow air into or out of the enclosure?

[Benta]

hi, depends.

[Someone]

As above, depends. Luckily those choices are easy to test:
https://sound-au.com/heatsinks.htm#s18

[pete g]

i meant blow air into or suck air out of the enclosure. thanks someone for your response. i read the article you proposed. blow in, is the choice.

[james_s]

In theory it doesn't matter, any air you blow out has to come in from somewhere and any air you blow in has to go out somewhere. In practice how effective either is depends on the layout of what's in the enclosure and where the heat is being produced. Sometimes blowing air in allows you to blow cooler air directly at some heat producing parts. It's possible to design an enclosure to work well with either arrangement though.

[srb1954]

It is often better to blow air into the enclosure so, with the appropriate placement of the fan, you can direct the highest velocity air flow to a hot spot for best cooling. Another advantage of blowing air into the enclosure is that you can fit a filter to the fan and reduce the amount of dust drawn into the equipment.

Sucking air out of the enclosure draws dust in at any and all entry points. It used to be a problem with early PCs where they would draw dust-laden air in through the floppy disc drive door slots and that dust would get then deposited over the internals of the floppy disc drive.

[pete g]

thanks to all who responded. pete g

[macboy]

Consider that pulling air out will also rarefy (thin) the air in the enclosure. Pushing air into it has the opposite effect, making the air inside even denser. Denser air will cool better.

[james_s]

Consider that pulling air out will also rarefy (thin) the air in the enclosure. Pushing air into it has the opposite effect, making the air inside even denser. Denser air will cool better.

I wonder if that has a measurable effect. The static pressure produced by a typical cooling fan is pretty low, I don't think it is capable of compressing the air to any meaningful degree but perhaps it does have some small impact.

[Nominal Animal]

Consider that pulling air out will also rarefy (thin) the air in the enclosure. Pushing air into it has the opposite effect, making the air inside even denser. Denser air will cool better.

I wonder if that has a measurable effect. The static pressure produced by a typical cooling fan is pretty low, I don't think it is capable of compressing the air to any meaningful degree but perhaps it does have some small impact.

The pressure differential is minimal –– we're talking about less than 4 mmH₂O, i.e. less than 40 Pa, or 0.04% of the nominal atmospheric pressure of 1 bar, with typical quiet 120mm or larger fans.
At the upper part of the range, it is just enough to reduce the amount of dust collecting inside the enclosure a bit.

I've built a few PC enclosures myself, and the difference is just big enough so that you want to have more forced airflow (fans) blowing in than out.  If you arrange for turbulent airflow inside the enclosure using baffles, you can actually keep surfaces clean of dust.  For example, 140x140x25mm Noctua NF-A14 PWM can generate about 2mmH₂O pressure, i.e. 20 Pa over ambient, moving 80 CFM or 140 m³/h, while generating less than 25 dBA measured at 1 meter distance.  Bloody expensive, but among the best 140mm 12V fans at of May 2023.

Anything higher, and we're talking about fans that necessarily generate 40 dBA or more noise at 1m distance.

[Georgy.Moshkin]

blow out of enclosure is probably a better choice. It will prevent water droplets from being sucked in during rain or water leak, especially if fan is powerful.  Another thing I was taught in 90s for is that blowing in will increase of dust/dirt build up inside enclosure, e.g. insects. Blowing in will lower temperature by some amount, but if we reverse power supply fan we may end up with cooler power board, but hotter air inside the enclosure, because air will be warmed up by power board components first.

[eTobey]

If you would use (need) a filter, then you should blow in, with a filter on the fan. This way there will never be dirt in it.

[RoGeorge]

A fan with zero pressure on the sucking side would do nothing.  The lower the pressure on the sucking side, the worst.  Therefore, it would be preferred to position the fan such that it will pump the air into the enclosure, not out of it.

[Nominal Animal]

Another thing I was taught in 90s for is that blowing in will increase of dust/dirt build up inside enclosure, e.g. insects.

Practical experimentation has shown the opposite.

[PeteH]

Other thing to consider is the fan itself would be running at higher temperatures if it was pushing the heat out, vs being at the external air temperature.

[Miyuki]

Some 1U and 2U rack enclosures have fans in the middle
So they suck and blow at the same time 

[Nominal Animal]

Some 1U and 2U rack enclosures have fans in the middle
So they suck and blow at the same time 

Sure, all my PC enclosures do that too; consider the fans on the heatsinks inside the enclosure.  I also tend to have both intake and exhaust fans.

It's just that ensuring there is a little bit of overpressure, i.e. more intake than exhaust, and turbulent airflow over surfaces, seems to work best overall.