Table of Contents



Literature Review

Experimental Design


Test Station Construction





Statistical Analysis


Glossary of Terms




This was a really interesting project and I enjoyed working on it. I wanted to do a project that was important and relevant to society. Phase 1, the 2003 "Torque it Up!" project produced results that strongly supported the idea that a multiple rotor windmill design was potentially more efficient than conventional three blade windmill designs.

In Phase 2, "Watts Up with Torque", I wanted to further my research and use a more direct approach than torque, to measure windmill efficiency.

For the project "Watts Up with Torque!", I was originally going to use a pulley system connected to the windmill axis and the axle of the motor. I quickly discovered that this was not a viable option. The pulley kept slipping and when tightened, pulled the motor out of alignment. A visit to a local specialty model airplane shop proved successful. I was able to purchase couplers to connect the windmill axis to the motor axle as a direct drive. The couplers allowed me to connect and disconnect the DC motors relatively easily. This modification ensured that errors in measurement were minimized.

The use of couplers and various sizes of DC motors allowed me to not only measure the electricity generated, but also allowed me to assess the ability of a rotor variable to start and continuously turn the axle of a motor.

"Watts Up with Torque!" started out as an expansion of Phase 1. I wanted to test the effect of more than two rotors and the associated variables of placement, blade orientation and size on electricity generation. Choosing to test various sizes of DC motors provided the added bonus of verifying my previous results about torque.

This project allowed me to learn how to use additional devices such as the multimeter for measuring mAmps and mVolts and the DC motors as generators. The experiments taught me how careful I had to be in positioning the instruments and being patient enough to wait until the RPM of the rotors stabilized before attempting multimeter measurements.

In all, I planned to collect 2880 observations. The actual amount of data collected was less because some of the rotor variables failed to start and/or continuously rotate. I learned more about how to organize a controlled experiment, and design useful worksheets for recording the data collected. I learned a lot more about spreadsheet programs and how to use them to analyze and present the data collected. I also learned some basics about statistics and how they can be used to assess the quality of the testing performed.

My multiple rotor windmill design seems to be very efficient and worthwhile pursuing. I may continue to explore additional facets of this approach in future projects.

Sources of Error

The experiment was designed to keep sources of error to a minimum but not every aspect could be perfectly controlled.

1. Although the same three speed fan was used for all tests, the wind speed measurements at each speed were not exactly the same. However, the wind speed data at each speed was analyzed using statistics and there did not appear to be a significant variation.

2. Some of the rotor variations would start the motor turning, but would not allow it to run continuously. It was impossible in these cases to get consistent readings of RPM, mAmps and mVolts. In these cases, that rotor variable was assigned a RPM of zero.

3. The wind speed, RPM, mAmp and mVolt measurements could not be done simultaneously. The anemometer was held between the fan cage and rotor and the position interfered with the wind flow to the rotor, meaning that RPM, mAmp and mVolt measurements would not be valid.

4. The anemometer had a stated accuracy of +/- 3 % or +/- 0.1 m/s.

5. The tachometer had a stated accuracy of +/- 0.05 %.

6. The multimeter had a stated DC voltage accuracy of +/- 0.5 % and stated Amps accuracy of 1.2 %.