A silly thought concerning case fans

[Shrek]

Imagine a tube with fans at each end and compare this to the same tube but with the fans at the same end.

In the first case

• Series: air passes the first fan and then the second (after traversing the case)

while in the second case

• Parallel: the fans don't double process the air (they are side-by-side not stacked)

I wonder which is best?


If I was very hot and had two room fans, I would probably favor having two run in parallel (next to each other) rather than run them in series.

 

[freeagent]

With my old R4 I would run a helper fan to help draw the cool air from the front fans further into the case. It worked ok.

 

[evernessince]

If by on the same end you mean stacked right next to each other than that is going to be worse. When stacking two fans you are obstructing the first fan with the second, causing a lot of turbulence and deflection. You'll have air being pushed by the first fan bouncing off the back of the frame, hub, and blades of the second which will increase pressure against the first fan and recirculate air.

In addition two fans of the same model running at the same RPM close to each other tend to cause a whirling noise as the noise frequencies they produce interact. GN recently did a video on this topic.

There are fans that are industrial fans with multiple fan hubs and blades and those have to be designed specifically to account for all the above factors.

You'd make the config much better by spacing the fans out.

 

[JWNoctis]

If by on the same end you mean stacked right next to each other than that is going to be worse. When stacking two fans you are obstructing the first fan with the second, causing a lot of turbulence and deflection. You'll have air being pushed by the first fan bouncing off the back of the frame, hub, and blades of the second which will increase pressure against the first fan and recirculate air.

In addition two fans of the same model running at the same RPM close to each other tend to cause a whirling noise as the noise frequencies they produce interact. GN recently did a video on this topic.

There are fans that are industrial fans with multiple fan hubs and blades and those have to be designed specifically to account for all the above factors.

You'd make the config much better by spacing the fans out.

I -think- if one of the fans runs in reverse direction from the other and there's otherwise more or less free space in the direction of the airflow i.e. not into a heatsink fin stack, the airflow would likely be much greater due to some flow dynamics shenanigans if you stack them. I had a desk fan that utilized the effect. It was quite powerful for its size, and gave little airflow if I held down one of the two rotors with my thumb.

 

[Shrek]

If by on the same end you mean stacked right next to each other than that is going to be worse.

I meant next to each other, not stacked for the parallel configuration and at each end for the series configuration.

I clarified the original.

 

[pressing on]

I meant next to each other, not stacked for the parallel configuration and at each end for the series configuration.

As well as "intake" you need to consider "exhaust". So having one fan blowing at you and one fan blowing away from you might be better. Or having the second fan blowing outwards through an open door or window.

 

[Vario]

With a computer case you want the air inside the case to be completely exchanged with the air of the room, and you want the components inside the case to transfer their heat to the air inside the case, so that the heat makes its way outside the case when the case exchanges its air with the room air.

If your case only has several vents, having all the fans intake or having all the fans exhaust or having half the fans intake half the fans exhaust really doesn't make much difference. If the case is more ventilated, different fan configurations should be tried.

 

[Shrek]

An extreme example, but real; I have a gaming computer with 3 intake fans and one exhaust. It is possible that the one exhaust fan actually obstructs the flow.

 

[taka]

Paralel running extended blade fans are quite common in servers. This is a perfect design if you dont mind the noise 15K RPM ! 12v 1.8Ah
Here is a Dell 730xd.

Fan's suck air from front grill cooling the drives and BMC, forced over Ram and CPU heatsinks and all that's behind (slots, PERC, iDrac, etc.)

Here is the full assy with air ducts

moving the fans at one end as intake or exhaust i guess it's the same airflow, more or less...

What you can do is experiment with a cardboard box and fans.

 

[Vario]

An extreme example, but real; I have a gaming computer with 3 intake fans and one exhaust. It is possible that the one exhaust fan actually obstructs the flow.

It could if the exhaust fan was really weak and the intake fans were really strong, or if the intake fans were really big and the exhaust fan was really small. Otherwise it doesn't matter much, air is going to find its way out. Ideally you do want laminar airflow so that the air enters the case, goes straight out, minimal turbulence.

 

[Shrek]

I am overthinking this whole thing... but it sure beats work .

However, I am not sure about avoiding turbulence as turbulence helps with heat transfer.

Parallel running extended blade fans are quite common in servers.

Parallel makes sense to me, that is exactly why I was questioning series.

 

[freeagent]

I like turbulence.

 

[taka]

Parallel makes sense to me, that is exactly why I was questioning series.

Series also make sense in some scenarios but with equal push-pull config.

Remember the Antminer?

 

[Shrek]

I could see a use for series fans, namely redundancy so if one fan fails, the other is still running.

 

[JWNoctis]

It could if the exhaust fan was really weak and the intake fans were really strong, or if the intake fans were really big and the exhaust fan was really small. Otherwise it doesn't matter much, air is going to find its way out. Ideally you do want laminar airflow so that the air enters the case, goes straight out, minimal turbulence.

I doubt there'd be much laminar airflow inside a case, downstream of a spinning fan blade into a case with cables and VRM heatsinks jutting out, a graphic card, and, if using air cooling, an often rather huge CPU heatsink squat in the middle. It's not like we are optimizing for pressure drop, and mostly it's just the amount of airflow that mattered. Mixing and minimized stagnation is also useful. The objective is to get the heat out of the case as quickly as possible.

An extreme example, but real; I have a gaming computer with 3 intake fans and one exhaust. It is possible that the one exhaust fan actually obstructs the flow.

If you have a dust filter on those intake fans, your computer's insides should also get dusty much slower. Normally the exhaust fan should not obstruct the flow, unless you leave it off.

 

[joemama]

I would guess it depends how much resistance is there to the air flow, if there are lots of obstruction within the case, then one fan on each end would be the best, otherwise, two fans in parallel would be better.

 

[micropage7]

push pull giving you better performance

 

[lilhasselhoffer]

You have a "series" versus "parallel" question.

In the series fans (push-pull), the air is accelerated by one into the case, and out of the case by another. The benefit to this is that if the fans are equally specified you should have a flow-through which is consistent (roughly), and somewhat more laminar flow....though that's a huge assumption that you aren't just blowing right into a fin block.

In the parallel case you have to assume that the accelerated air diffuses from whatever opening there is...IE an opening or poor sealing. A fan is functionally incapable of pumping air...so all you can do is push it and let it naturally diffuse.


If you want the most flow over components you want series. If you wanted to be cool, and we assume this is a "you the squishy human," then increased flow allows for greater evaporation of water (sweat), which is your primary mode of cooling. We would choose a parallel situation if we needed to push something through a filter, and then didn't care how it diffused. This is often times seen on computers with a front panel dust filter...and it takes advantage of creating a huge fouling layer on the filter to clean out the air, before being allowed to naturally diffuse inside your case and cool through standard flow. This is why most cases seem to have only a single output in the back...but 2-3 minimum fans in the front. It's insanely useful to not have excessive dust build-up.

 

[Athlonite]

An extreme example, but real; I have a gaming computer with 3 intake fans and one exhaust. It is possible that the one exhaust fan actually obstructs the flow.

This makes for positive pressure inside the case and if you had them the other way round 1 intake and 3 exhaust you'd then end up with negative pressure but having 2 of each would equalize the pressure in the case and cause the air in the case to be removed faster as the air being bought in would also be faster

 

[Caring1]

This makes for positive pressure inside the case and if you had them the other way round 1 intake and 3 exhaust you'd then end up with negative pressure but having 2 of each would equalize the pressure in the case and cause the air in the case to be removed faster as the air being bought in would also be faster

Depends on the case, it it's a fish tank or wholly mesh.

 

[Shrek]

push pull giving you better performance

How so? one is comparing two feeds against one.