Fan Shrouds - Does depth make a difference?

[KuniD]

Does the depth of a fan shroud make a difference to how a fan/radiator combo performs?

1" depth vs 1 5/8"


If it doesn't make much of a difference I'd rather go for the 1" to save space.

[bobkoure]

Caveat: I watercool pretty much for low-noise. I tend to dial down fans until they're unobtrusive (guessing under 18dB) and then look at MBM-reported temps, both idle and CpuBurn. I do overclock, but low-noise gets priority.
I've found, in my very limited experimentation, that plenum size (shroud depth, if you prefer) makes more difference if you're pushing air through a radiator, rather than pulling - so long as the "pull" side of the blades has no obstruction within, say, half a chord-span (otherwise you get interference with the fan - they're not so much augurs as wings so coanda effect is important - and can be easily screwed up).
I've also found that you can't hurt performance with too much plenum, and you may find additional benefits. For instance, if you use plenum on the pull side (where more doesn't seem to help temps), more does seem to reduce noise a bit.
Of course, this only applies to the extent that what you're doing matches what I'm doing - and your question was pretty telegraphic...

[Jimbo Mahoney]

What's half a chord-span bob?

I'm guessing half a blade length?

[KuniD]

Just had a quick re-read of pHaestus' WC tips, I remembered he had linked to a good article on shrouds.

http://www.nmbtc.com/pdf/forum/engineering_101.pdf

So if I understand this right, I need to keep a certain amount of space on both sides clear.

From reading various threads here I get the impression that pulling through a rad is better than pushing air into one. The rad is going to be placeed right at front of the 5.25" bays on my case.

Depth in brackets

fresh air (casefront wall) |> RAD(1"/25mm) >shroud(1"/25mm)>Fan(1"/25m)>Shroud(1"/25mm)

So the fan is sucking air through the rad from the outside, and its got a 1" shroud on both sides.

This takes up roughly 4"/10cm from the front of the case, leaving another 8-10cm behind for me to mount two cylinder shaped reservoirs (6cm diameter.) Hopefully these won't mess with the airflow too much.

[bobkoure]

Originally Posted by Jimbo Mahoney

half a chord-span?

Oh, sorry - mu co-workers are a bunch of airplane and soaring geeks and I forgot it's specialized knowledge.
The blade of an axial fan is a wing. A chord is an imaginary line connecting the leading edge with the trailing edge. Coanda effect is, essentially, progressive laminar flow across the top surface of the wing that turns it into a scoop, pushing air down behind the trailing edge. The effect extends upwards from the wing, but usually no more than a half chord (depends on the wing, airspeed, some other stuff which I mostly don't understand), but it's a top surface effect. Ever notice how most axial fans are exhaust-over-struts, or, for that matter, how fighter planes bolt all sorts of stuff to the underside of the wing?
Anyway, for the axial fans we mostly use, it isn't much distance at all,...

[maxSaleen]

Just thought I'd mention that any shroud is better than no shroud at all. Will there be a difference between 1' and 1 5/8'? Probably. Which will perform better? I couldn't tell you. It would probably depend on your fans. Think practically here. If you only give up .5c and you gain 5/8' of case space is it worth it? For me it would be. The noise difference might be more noticeable, though there are more effective ways of dealing with noise than shrouds.

[joeyek]

Originally Posted by bobkoure

Caveat: I watercool pretty much for low-noise. I tend to dial down fans until they're unobtrusive (guessing under 18dB) and then look at MBM-reported temps, both idle and CpuBurn. I do overclock, but low-noise gets priority....

aha, me2, cause watercooling won't bring me 2 much advanced overclocking space above aircooling, 1 or 2 degrees hotter would be worth to pay for the lower noise.

Originally Posted by bobkoure

I've found, in my very limited experimentation, that plenum size (shroud depth, if you prefer) makes more difference if you're pushing air through a radiator, rather than pulling - so long as the "pull" side of the blades has no obstruction within, say, half a chord-span (otherwise you get interference with the fan - they're not so much augurs as wings so coanda effect is important - and can be easily screwed up).....

Did you mean the deep shroud is more effective in a fans' pushing setup than the pulling one? When the factor of the rebounded air by the rad is taken into account, the situation could be different. More space can avoid the dead zone of the axial fan, but at the same time the force is weaker. Like aircooling, a weak fan cannt blowing the air deep enough to the hottest base of a heatsink.

[Jimbo Mahoney]

Originally Posted by bobkoure

A chord is an imaginary line connecting the leading edge with the trailing edge....
Ever notice how most axial fans are exhaust-over-struts, or, for that matter, how fighter planes bolt all sorts of stuff to the underside of the wing?

Cool, thanks dude.

I think my 2" shroud should do for my 120mm fans then.

[bobkoure]

Originally Posted by joeyek

Did you mean the deep shroud is more effective in a fans' pushing setup than the pulling one?

Yes, that's what I observed.

When the factor of the rebounded air by the rad is taken into account, the situation could be different.

I only have a few data points. Absolutely, the situation could be different if you are doing something different than I was - for instance, if you have a more-powerful fan, although I suspect (as in guess) that it would actually be more pronounced in this case.

More space can avoid the dead zone of the axial fan, but at the same time the force is weaker.

I don't presume to understand exactly what the air movement is doing. It'd be interesting to get an array of different rads, fans, and clear shrouds and use smoke to see what was actually happening. Failing that kind of data I don't know what's going on in there - and I'd suspect you don't either.

Like aircooling, a weak fan cannt blowing the air deep enough to the hottest base of a heatsink.

This is actually quite a different situation and one I do know a bit about as I did a certain amount of race bike work in the 70's. Air cooling a heat sink is very like air-cooled motor cooling in that you are trying to force airflow down between fins and then out to one side or the other. Directed pressure is important here as you otherwise end up with a standing wave (higher pressure, if you will) at the root of the fin and air only penetrates to the point where the two forces balance. (And yes, we did testing with smoke - come to think of it, there were some mirrors involved as well to allow simultaneous filming of several areas, so it really was "smoke and mirrors" ) Anyway, without appropriate ducting (including ducting an exit path to a low pressure area) longer, tighter fins, often could mean a hotter motor. When I got to talking to one of my aviation friends about this, he explained that this was very old news, so it's not like we discovered any new principles or anything.

[Ruiner]

BillA answered a similar question of mine here:
http://forums.procooling.com/vbb/showthread.php?t=11012

[KuniD]

How does this look?



Thats a top down view (to scale) of my 5.25" bays.
(http://www.kunid.com/projectdensity/1.03.jpg << picture of bays.)

The black block on the far right is the radiator, mounted right at the front of the 5.25" bays, then next on the left is a 1" shroud, then a 120mm fan (also 1" depth), then another 1" shroud on the opposite side, so the reservoirs (adapted version of Trodas' Tank o Matic) don't get to close. The big red block in the background is a Pioneer DVDRW drive which'll be right at the bottom of the 5.25" bays.

[Jimbo Mahoney]

I'd be intrigued to know if the exit shroud has any / much impact on temps / noise.

If it does, I'll mod my shroud / fan setup.

In fact, if, in theory, it's supposed to help, I'll bodge one together and see anyway.

[bobkoure]

Can you "stack" the shrouds? If so, You're set up nicely to see if changing between 1" and 2" makes a difference.
My guess is that it won't make a difference temp-wise, but will be a bit quieter. If there's room to put foam on the inside of the shrouds without obstructing the airway, that should quieten things a bit. (I'm thinking of the really thin stuff like Akasa, which is only 4 or 5mm thick - so it'll only attenuate the very high frequencies, but those are the irritating ones, right?

[maxSaleen]

The very high frequencies are the irritating ones but the cavet here is that we are using 12cm fans which tend to put out lower frequency noise. My personal feeling is that thicker material is needed than akasa to provide any respectable sound absorbtion. I've used it before and did not notice a substantial reduction in noise.

[bobkoure]

Absolutely. Absorbtive materials only work to the extent that there's some sort of restriction a half-wavelength out from whatever the sound waves are reflecting off of - and this foam is really thin. On the other hand, it's cheap, easy to apply to already-built shrouds and will probably do at least something - just not a whole lot.
I've done it with a couple of shrouds. I'm not sure the noise level went down (I think it did but it was small enough a difference that I could easily be imagining it), but the MBM readings stayed the same, so it didn't seem to hurt temps, or at least not enough to get noticed by MBM.