Estimating airflow through a ducted fan/heatsink

[pHaestus]

This very question has been coming up as Joe and I kick around ideas for testing heatsinks:

What is the airflow through the actual heatsink, and how can it be measured?

http://www.electronics-cooling.com/h...august_a3.html

"Under these conditions, measurement of the coolant flow rate is problematic since insertion of an obstruction meter, such as a laminar flow element, into the fan/heat sink test article coolant flow path will introduce an extra flow resistance, so that the operating point of the test article will be altered. This difficulty can be circumvented by using the test article fan, in conjunction with its performance curve, to determine the flow rate through the system. Only two easy measurements are required: the fan speed and the pressure rise across the fan."

Now to get appropriate pressure drop measurement gear....

[bigben2k]

(my post dissapeared!)
I've often thought about that. I figured that a manometer, similar to what has been previously suggested to measure pump pressure differential should work but I can't figure out the placement of the opening.

1-if the opening is facing the fan, then there would be an error because of higher than actual pressure

2-if the opening is facing the other way, the pressure would be lower.

3-if the opening is perpendicular, then there would be a restriction, from the crossflow.

I suppose that a perpendicular opening that is fairly large should do it. I dunno.

[bigben2k]

What I didn't mention is that one of those (#1 or #2) should cancel itself out on the backside.

Here's a few more links:
http://www.govictory.com/How%20To/FL...FLOWBENCH.html

http://sanyodb.colle.co.jp/product_d...pdf/e_spec.pdf

http://www.specialinstruments.com/pdf_e/28-29.pdf


It then becomes apparent that since air is a compressible fluid, the measurement may do well to include this as a factor.

[utabintarbo]

How do electronic fuel injectors measure mass air flow? I think they measure the pressure drop. Does this help, or should I go back to lurking?

Bob

[BillA]

Originally posted by pHaestus
. . . . . Only two easy measurements are required: the fan speed and the pressure rise across the fan."

Now to get appropriate pressure drop measurement gear....

well, as usual it ain't really all that easy
"fan speed" - no sweat, optical tach
"pressure rise" - hmmmm . . . .

the equipment is no biggie (hey, its only money), a digital differential manometer can provide resolution to 0.1 Pa
(less pressure than one wants to even try to imagine)
(and those Dwyer magnahelics are WORTHLESS)

so then the "pressure rise" would be ??
- if its drawing from 'free air' then the low side is atmospheric pressure, easy, leave one side open
- and the high side is . . . .

slight problem here, the fan is slapped up against the fins
- so there is certainly some uneven pressure 'distribution'
- and where and how to measure ?
-> understand that there is a venturi effect across the end of a tube that becomes very complex very quickly in terms of 'calibration'

BTW, my rad testing used 12 to 15+D duct with an open discharge to sidestep this
(extremely flush opening in sidewall - no tube)

keep this alive pHaestus, I'm interested how you and Joe resolve this

[Bignuts]

Will a manometer help me get fitted for a ProGear man-bra?

[pHaestus]

Bill: Would you think a 1-2" shroud between the heatsink top and fan could (mostly) deal with this? I don't see why that would be such a big deal to implement. People really should use such anyway with fans.

[BillA]

I would agree that hsfs 'should' have such - but they don't
but if all were tested that way . . . .

I guess the question might be resolved by testing one with, and then without
and looking (with some precision) to see if there is a difference (a difference of significance ?)

I have a digital manometer and 0.01° resolution (blah. blah)
BUT no hold down lugs

want a loan of the manometer ?

[Brad]

how about getting a rubbish bag (of a measured volume), or something similar. Seal the opening over the top of the fan - which is mounted on the heatsink. Turn the fan on, and measure the time it takes to suck all the air out.

Not the most accurate way, but if someone tried to do that 5 times in a row and got repeatable results then it might be worth a look into

[g.l.amour]

dunno if u guys know the expression 'the egg of columbus' but brad seems to have found it anyway. i should prolly just shut up...

[BillA]

not familiar with "the egg of columbus", do explain

I was going to suggest that he use different colored smoke bombs,
capturing the output of each in separate bags (first evacuated by sucking on them),
then he could measure the output of each color; 1 1/2 bags of red smoke, 1 3/4 green, and . . .

holy sh*t - 2 1/2 white; look how much faster the white smoke is !!

[UNDERBYTE]

Hotwire anenometer-measure the velocity of the exit air(on a grid and average the result), calc the exit area and you have enough info to to calculate flow through the sink

or

Build your own test chamber -AMCA calibrated nozzles can be bought for a few hundred dollars

some plywoood or plexiglass and and a counter blower. total cost about $1000 USD

many fan co. build thier own

Measuring airflow is not going to tell you much of anything useful unless you are doing development. your constant should be @ equal noise values vs performance. Puts every thing on a level playing field....

[BillA]

I use a Hotwire anenometer in a duct, gridding etc
but I think the measurement of the exit air velocity off an hsf would be quite difficult
(and then to validate, whew)

your suggestion UNDERBYTE of "noise values vs performance" is excellent
as those are the parameters of direct relevance to the user

pHaestus
how different are the fan only noise levels and the fan plus hsf ?

from a hsf design perspective the airflow is needed, but not really by the users

I have already done the same thing for wbs, plotting "C/W" vs. pressure drop
very informative