measuring air speed

I am trying to come up with a way of measuring air speed in m/s for use in a home built wind tunnel for RC airplanes. What I have come up with sounds a bit involved (structurally) and I am wondering if there is a better way or if someone has done this before.

My idea....

use a 3 blade prop in reverse (9 inch diameter) coupled to a shaft with a disk of known circumference (fraction of a meter) with a single window notch in the solid disk. For example, one rev would represent circumference travel of 8 inches. So I would count pulses over 1 sec time period and multiply by a constant of proportionality to scale it up to meters/sec.

Am I over simplifying it or over complicating it? LOL....all I want to do is wing lift / drag tests on custom wing designs in a wind tunnel. BTW, the fan is an industrial movie fan with variable speed.

Best Regards,
Nick

Well being as I have experience in fluid measurement here is an idea. May be invovled but it is used in high precision applications.

Air is a fluid. As the circumference of a tunnel decreases, the speed of the fluid increases until it reaches a supersonic velocity then the inverse applies. So construct your wind tunnel so that the diameter decreases after the wing that is being tested. Take a pressure measurement before the decrease in size and after the decrease in size. The change in pressure (should go down as velocity goes up) is directly proportional to the velocity. Pressure can be taken with pressure transducers. You would then multiply the change in pressure by a constant and that will give you the velocity. I will calculate the constant if you should choose to use this as it can be complicated and is based on the change in size. Hope this helps.

Travin

Here is an interesting site for you to read http://www.grc.nasa.gov/WWW/K-12/airplane/pitot.html.

Hi, MacG.

To have a correct measurement ... Travin is the closest to the truth !!!

This is called Venturi effect, and it's used in Big wind tunnels. Pitot tubes would be too false here, due to the non isotropic flow.

But, before entering the Venturi section you must have to redirect air flow, cause of the disturbance created by your wing section, that would affect the mesurements. ( we can't suppose fluid is uncompressible here ... )

Remember at low Reynolds ( say ... 50k to 1M ), flow disturbances are close to pressures you measure.... you will be very far from sonic flow !!!

Back to the basic subject ... Your PC certainly is the most precise virtual wind tunnel you could have ... and there are lots of really good free software on the net !!!

http://www.designfoil.com/

Alain

Hi Nick,

Wow, a real wind tunnel in your workshop. That must be the sign of a serious hobbyist!

Knowing nothing about wind tunnels, I am not sure if this idea would work in the m/s range.
There is a method, similar to the "hot wire anemometer", to measure the flow of air.

Using "King’s Law", to monitor the cooling on two "self heated" transistors. The transistors also allow the monitoring of it's own temperature (built in), and regulate the self heating currents. I have seen a couple of short articles (by: W. Stephen Woodward), with schematics. If this sound interesting I will look for the references.

Thanks for an interesting problem,
-Adam-

If I am not mistaken, bernouli's principle states that as flows across a wing, lift is generated based on the fact that as air flows over the top, it has to flow faster which makes the psi go down and creates lift. That being said, the speed of the air at the rear of the wing should equal the speed at the front of the wing. I think that any difference in speed is going to be from inertia of the air over the top which is based on the mass of the air which should be pretty damn small. As long as the venturi nozzle is about 12 - 24 inches away and is free flowing, you won't get a significant back pressure. If you need help with the design let me know and I will see what I can whip up.

Travin

Thanks all, your ideas have gotten my creative juices flowing. I need to think about how to set the meter structure up. Thanks again!!!

Nick

I have designed PIC-based air speed sensors. The method that I usually use is to take a PTC (positive temperature coefficient) thermistor with a transition temperature of about 120C and apply 12V across it. The thermistor acts like a heater and thermostat all in one, and will consume just enough current to maintain its temperature at the transition temp. The cooling effect of air flowing over it depends on 1. The square root of the air velocity (King's law) and 2. The ambient temperature.

When selecting the PTC thermistor, you have to be certain that the device is a very small one and has low enough resistance to achieve its transition temperature with only 12V applied across it.

I put a small resistor (4.7 ohms is a good value) in series with the PTC thermistor and amplify the voltage drop across the resistor with an op-amp. I feed the output of the op-amp into an A/D input of my PIC.

I use a normal negative-temperature coefficient thermistor in series with a resistor to measure ambient temperature.

I can then calculate the air velocity in feet-per-second.


Charles Linquist