How to test CFM of various fans

[NHerron]

Anyone know of a somewhat simple way to measure airflow of various radiator fans?

I saw this thread:
Electric Cooling Fan CFM Rating Tests - DirtBound Offroad - NAXJA Forums -::- North American XJ Association

I could swing the money on one of those meters (http://inspectusa.com/mannix/DCFM8906.pdf)
but I'm a bit more anal about accuracy.

Then I thought about using something like a GM airflow sensor (with a makeshift ducting cone tunnel thing) but don't think they measure high enough according to ECMtuning:
maffrequencylimits [ECMTuning - wiki]



I'm trying to figure out what brands are stretching their specs (not trying to harsh on any brand however- just wanting to get the best flow)

 

[92nolatist]

Im not sure about the whole gm maf thing. But I think you can use an anemometer (wind meter) to measure cfm. I did a quick google search, 'using an anemometer to measure cfm' and yielded a bunch of results. Those do get expensive though, so youd have to stick with a cheaper model if you dont want to put too much time and money into it.

http://www.google.com/#safe=off&q=a...16,d.dmQ&fp=27dc0fd669a0b71e&biw=1280&bih=628

 

[____Kurt____]

Wow those sure are expensive. For the prices you can buy 2 spal fans and call it a day and rest assured you got some good flow if not the best. OEM should always be first pick but they are just too big. Spals should always be second choice. I got mishimoto slims and they are pretty lousy.

 

[NHerron]

Did a little reading

Found out the GM MAF sensor is an Anemometer in a round-about-way. Check this article out:
HowStuffWorks "Anemometer"

How about that, a hot-wire anometer

And the propeller anemometer's are accurate, I just wasn't understanding how it is measured.

Basically you punch in the parameters it asks for such as diameter, shape etc. Then you take a bunch of samples and average out the results. If anyone knows transistors its somewhat the same idea, based on the input parameters and the sample airflow, it boosts the sample signal to get the big picture.

I'll shop around for the best one and see how it goes.

I'm happy, thanks

 

[91-GS]

If you are looking to compare fans (as in "this one blows more air than that one") you can take a small electric motor (like out of a toy car) and mount a propeller on it's shaft. Connect a volt meter to the motor and place it in the air stream coming off the fan. The volt meter will give you some reading, and higher the CFM of the fan, higher that reading will be (since the little propeller will turn faster and spin the motor that is now a generator faster, making more electricity).

Place the rig in a stream of a fan with known CFM and see what the volt reading it. From there it will be pretty easy to calculate actual CFM of other fans. Would also help to have the fans in a duct of some sort to keep air from going "out". And do keep the propeller (the DIY anemometer) same distance away from fans for consistant measurements.

 

[NHerron]

I like DIY projects for sure (especially electronics), but I have no time for experimenting when working 13hr days. Wish I did.

Besides, I was thinking about posting the results so the info should be the most accurate.

 

[tkelly27]

A hot wire air sensor essentially measures the speed of the air in a given area. If you put the element into a larger area of pipe you will increase the amount of airflow it can read. This is why you need to re-calibrate for the 3.5" MAF if you've been using a 3" MAF.

The change in air flow is based directly on the change in area, which is based directly on the square of the radius of the pipe size. For example, going from a 1" pipe to a 2" pipe means that you are flowing ((1^2)/(.5^2)) = 4 times the amount of air. Yes, to find area from radius you need to use pi, but that is a constant that gets dropped when you divide. Try it for yourself here

Area of a circle calculator

So you can easily find the size of the housing you would want to use by finding the ratio of the airflow you want to have (3000CFM should be enough to test a fan) to the max airflow of whichever sensor you use. That will be the same as the ratio of areas, find the unknown area, which gives you the unknown radius, which gives you the unknown diameter. Easy.

Some sensors are removable from their intake pipes. Ford and Subaru are ones I can think of off the top of my head. Being a DSMer you may have a DSM ECU and a MAFT handy. That will make power supply and datalogging really easy compared to other platforms.

 

[PieEyedPiper]

Using the gm maf idea you might find yourself to be testing some measure of static pressure rather than the cfm. IMO, there are 2 important aspects to a fan: cfm, static pressure and noise. OK three.
But 'eff noise in this situation. The restriction from nozzling down from say 20" to 3" is really extreme. It's possible that this kind of restriction could reduce cfm to some meaningless number, whereas, a 3 blade, heavy pitch fan design might produce similar numbers to the cfm rating under that type of restriction.

I guess what I'm trying to say that this will certainly be an experiment and there's certainly a lot to consider.

Subscribed.

 

[daren_p]

We do airflow at work, it basically depends on how accurate you want to be. All the instruments posted in this thread would go under the "junk" category. Pay close attention to how their spec'd, it would be like amplifier power ratings, lots of different ways to fudge the results.

Both thermo anemometer & rotating vane would work for this. You would want to install the fan in some sort of duct & measure by what is called a duct traverse (basically an average as flow will not be identical in every spot). A lot of the cheeper units, up to the few hundred $ range are not very accurate. For example they may be rated 3% of full scale, if the instrument reads 0-6000fpm, it can read +/- 180 at any point & still be considered in spec (plus there's usually a fixed value added ontop, as well). Ontop of that, a lot of the time they are basically lying & even when new, they won't meet their published spec. If you want a better meter, your looking about the $600 range to get something half decent. To compare a meter like this might be 3% of reading (not full scale) a huge difference.

There is one other way & that would be with a micromanometer with a pitot tube. Basically measuring differential pressure (pitot tube has two measuring points, to give you pressure of the flow & static), which can be converted to airflow by calculations or automatically on some of the better micromanometers. This is probably the cheeper route but depending on what kind of flow your looking at, they may not suite the application. They are very un accurate at low flow levels because the relationship between pressure & flow is not linear. At higher velocities, they work quite well.

On the other hand, you could just get a SPAL or Derale fan & know they are of good quality