A technical fan question with real consequences

Felger Carbon · Post by **Felger Carbon** » Wed May 16, 2007 1:24 pm

Take a round tube, 120mm dia, a foot long. Mount a Yate Loon D12SL-12 fan in one end. Run it at 12V and measure the air velocity at the other end, as the air leaves the round tube. If you know the CFM you can calculate the velocity. If you know the velocity you can calculate the CFM. But right now, the velocity is the question; when we have that we can move on to part 2.

Part 2: plug the far end of the tube. Cut a 1" diameter hole in the middle of the plug. Apply 12V to the YL fan and measure the velocity of the air coming out of the 1" hole. Question: is this velocity greater than the former velocity with the unplugged opening? Or is there some law of fan-physics that limits the velocity thru constricted openings, which is what we have here?

I dunno, which is why I ask the question.

But I know for a fact that any PC case with a 360mm side fan is gonna have comparatively constricted exhaust openings. That's not a bad thing, if the exhaust air can be moving faster than the inlet air...

jaganath · Post by **jaganath** » Wed May 16, 2007 3:53 pm

i love these questions; can you tell?

part 1 should be easy; use an anemometer,hot-wire, axial or any other type.

part 2: by plugging the tube you have introduced impedance; this impedance will be inversely proportional to the size of the hole. the fan has to develop static pressure to overcome the impedance and so flow will drop. so the overall CFM will be less than in the first case. however since from conservation of mass we know the mass flow rate is constant if you restrict the area velocity must increase (mass flow rate = density * velocity * area).

of course we could simply measure it, but that would be no fun. I think you would need the P-Q curve for the D12SL-12 to know the CFM at whatever the static pressure is inside the tube, and then plug it into the above equation to get the velocity at the nozzle.

is there some law of fan-physics that limits the velocity thru constricted openings, which is what we have here?

yes, choked flow, but I don't think we actually have that here.

some refs:
http://www.actrx.com.tw/technical_pq.html
http://www.orientalmotor.de/de/uploads/ ... 152222.pdf

jaganath · Post by **jaganath** » Fri May 18, 2007 12:44 pm

wow, i think i even bored myself there; still,not interesting enough to warrant a response, even just to say "you're wrong"?

cpemma · Post by **cpemma** » Fri May 18, 2007 2:46 pm

jaganath wrote:wow, i think i even bored myself there; still,not interesting enough to warrant a response, even just to say "you're wrong"?

I'm still reading the small print.

I think you would need the P-Q curve for the D12SL-12

And I can't find one here.

DryFire · Post by **DryFire** » Fri May 18, 2007 5:41 pm

I would be surprised there are many axial computer fans that could create conditions necessary for choked flow; certainly none that we're interested in.

Instead of an arbitrary hole, why not mount the fan to a heatsink and measure the difference at the end of the tube.

This would probably work best with tower style heatsinks.

jaganath · Post by **jaganath** » Sat May 19, 2007 12:29 am

And I can't find one here.

yeah,thats sucks;AFAIK there is no way to derive the PQ curve from first principles,it has to be measured; I don't suppose anyone feels like building an AMCA 210 (85) measuring chamber and chucking a spare Yate Loon in there? and can I also have the moon on a stick if we're giving out wishes?

cpemma · Post by **cpemma** » Sat May 19, 2007 5:26 am

jaganath wrote:AFAIK there is no way to derive the PQ curve from first principles,it has to be measured...

I wonder how far off you'd be taking an existing curve for a same size fan, marking 'max flow, zero pressure' and 'max pressure, zero flow' points for the DUT, and tracing an interpolated curve?

This is following my hypothesis that good engineering design comes up with the same solutions.

Though I believe Chinese saws cut on the opposite stroke to Western saws...

Felger Carbon · Post by **Felger Carbon** » Sat May 19, 2007 6:09 am

cpemma wrote:I wonder how far off you'd be taking an existing curve for a same size fan, marking 'max flow, zero pressure' and 'max pressure, zero flow' points for the DUT, and tracing an interpolated curve?

Panasonic's published curves have very different shapes. The PQ curves involve CFM measurements, and as someone associated with SPCR has pointed out, measuring CFM is just a tad tricky...

Me, I'm waiting for a $29.95 5-digit Extech CFM meter. Just point the meter at the fan, press a button, and there's yer CFM readout!

Post by **Rusty075** » Sat May 19, 2007 10:01 am

Felger Carbon wrote:The PQ curves involve CFM measurements, and as someone associated with SPCR has pointed out, measuring CFM is just a tad tricky...

Actually, no. PQ curves are generated with the ANSI 210 testing standard, andat no point in that do you ever measure CFM. CFM measures are, as we've seen, nearly impossible to get accurately, so the ANSI standard instead derives the flow rates from pressure differentials and nozzle geometries.

...which is why the SPCR fan testbed is not an ANSI 210-99 test chamber. (yet

)

Here(pdf) is a pretty good explanation for how PQ curves are generated with an ANSI 210 system. You'll find it very interesting.