How to test the power output on a wind turbine?

I'm on a team prototyping a water filtration system powered by a Savonius wind turbine; the energy from the spinning turbine is mechanically transferred to a reverse osmosis filter/pump setup. We're doing tests of various blade configurations and are trying to determine the power output by hooking up a small DC motor (1-3V, 5000+ RPM) to the shaft. We're seeing only mW outputs though and were expecting something significantly larger. How do we pick a motor for the job? Do we need to match the RPMs of the motor to the expected RPMs of the shaft (~150)? Any other ideas for testing power? Thanks for any help!

Picture of How to test the power output on a wind turbine?
sort by: active | newest | oldest
1-10 of 14Next »
sam D7 years ago
Maybe overlap the two buckets for better efficiency?

Cheeck out my build's design at:

When they are overlapped I think it helps the air moving in the 'pushed' bucket to exit the concave shape.

jeff-o8 years ago
Perhaps not so cheap to buy, but a bicycle hub motor is designed to operate at that rpm range. Get one that is brushed (instead of brushless) and you can use it as a generator. Make sure you give it a bit of a load or you won't be able to properly test its current output.
lemonie8 years ago
Like other comments you need more load on this. A 12v motor hooked-up to a suitable load might do better than what you've got at the moment. Perhaps you could get a car-cigarette-lighter and a radiator fan-motor for some heavy load? (I can't quite judge how big this is)
But that might be a bit much without strong wind, a large-ish RC car motor perhaps?

. An automotive alternator/generator would be an excellent choice. You can get them in sizes from ~50A (~600W) to >200A (>2400W). You will need to use a speed multiplier (pulleys?) to get the alternator speed above ~1500RPM.
moisture (author)  NachoMahma8 years ago
Speed-multiplier..yeah ok, that makes sense. Thanks. The automotive generators are going to be a bit too much for our tiny turbine, I believe. Something similar though may be possible.
. How much mechanical load the generator puts on you turbine is dependent on the electrical load (resistor bank) on the generator . If the generator sees a very high resistance, it won't produce much mechanical load. Hook up several high-power resistors in parallel with switches in series with each resistor. No switches on (infinite resistance) = low mechanical load. The more switches you turn on, the greater the load.
. V-belts are usually pretty stiff and may present too much resistance. I'd try a more flexible serpentine belt.
. A clutch may help in getting the turbine up to speed before applying any load.
moisture (author)  lemonie8 years ago
The turbines we're testing are only about 1.5 feet tall. I have a PC fan that I may try; the RC motor's a good idea, thanks!
Yes, the motor speed has to be matched to the rotor speed, however low speed DC motors are uncommon.

Why not just rig couple of pulleys and have it lift a weight as it spins, and time how long it takes to lift the weight a distance. mgh/s (kg*m2/s3) is a watt.

and here's a power curve graph for various turbines
moisture (author)  Tool Using Animal8 years ago
The pulling-a-string approach was an idea our team tried. It's difficult to create something that engages the weight system when the turbine hits its stable rpm.
i think the overall time (including before it gets to full rpm) still roughly represents the full rpm
1-10 of 14Next »