The end of September saw a major milestone completed on the wind turbine project – we finished our wind tunnel testing at the NASA-Ames research center! The experience was fantastic and so far we have been blown away by the quality and quantity of data we collected. While it is still a bit too soon to publish our complete test results here (we still have a lot of analysis to do), I nevertheless wanted to get started by explaining our test set-up and posting a few photos from the tunnel.

Test Set-up

Since the beginning of this project we have focused on two vertical-axis turbine designs: Savonius and Lenz (see photos below). The major purpose behind our wind tunnel testing was to characterize the performance of these turbines, allowing us to select the most promising design and move forward with developing the alternator that will generate the electricity. In order to characterize the turbines, we needed to collect data on how they perform in different wind speeds while under different loads.

The end result of this testing will be a series of power curves (see chart below) describing the mechanical energy generated by the turbine as a function of wind speed and turbine speed (rpm). We also explored three different angles of attack for the turbine blades on both turbine designs to see how they impacted turbine performance.

The theory behind our turbine test is this simple equation:

Power = Torque x Rotational Speed

When the wind blows it hits the turbine, causing it to spin. In spinning, the turbine converts the energy contained in the wind into mechanical power. In order for us to figure out just how much wind energy is converted into mechanical power we needed to place the turbine in a flow of wind with a known speed and then apply a known torque load to the turbine while measuring the effect of the load on the turbine’s rotational speed. Referring to the equation above, we would supply a known torque and measure the rotational speed, thereby allowing us to calculate power.

The way we mechanically applied the torque load was relatively straightforward. We coupled a DC motor to the bottom of the turbine shaft and tried to turn the motor opposite the direction of the turbine rotation (see image below). Since the torque produced by a DC motor is a function of current, we could apply a known load to the turbine by applying a known current to the motor.

This test set-up, while elegant, nevertheless proved problematic as the motor we sized proved unable to resist the amount of torque the turbine was generating and we quickly overheated the motor. Another way of applying torque to the turbine was needed, fast. Thankfully, the NASA-Ames facility is full of advanced testing labs, many of them working on rotating equipment. The day after we identified our problem we had a torque cell in-hand and connected to the turbine shaft (see image below).

This torque cell contained a strain gage mounted on a shaft that output a voltage based on how much stress was applied to the shaft. One end of the shaft was attached to the turbine and the other end had a mountain bike disc brake attached to it that could apply a drag load that would slow the turbine and provide the strain that the gage would measure. This set-up allowed us to apply our known torque load as desired and the testing went on as planned. The only downside was that we were unable to maintain a steady torque and were forced to take our readings dynamically as the torque applied and turbine speed varied. This scenario was less than ideal, but we still managed to collect a copious amount of data that should allow us to compensate for any dynamic and inertial effects.

Ultimately we collected all the data we wanted and, at first blush, the results look great! My next blog post will focus on that data and our resulting analysis. Many thanks to Malcolm Knapp, Jeremy Kimmel, Sarah Felix, and Charlie Sellers who all devoted many days to the wind tunnel testing. Other Engineers Without Borders volunteers that played an important role are Jerry Pugh, Matt McLean, and Ann Torres. Finally, none of this would have happened without the help of Jose Navarette, Nili Gold, and Farid (all of whom work at the NASA-Ames facility), the technicians running the tunnel, and the generous donation of the facility by the US Army Aeroflightdynamics Directorate (which leases the tunnel from NASA).

To see more images of our wind tunnel tests, check out our Wind Tunnel album and watch the video below:

The “holy grail” as it were of developing world solar power proponents is $1/watt. This is often seen as the point at which local electricity generation becomes affordable for the majority of the world’s population that do not currently have regular, reliable access to a regional electrical grid. This then begs the question, “Should $1/watt be the target of any new, small-scale renewable energy project?” Furthermore, if that price point appears unattainable is it foolhardy to chase after a technology that is otherwise very appealing? The only way we can know for certain is by dealing directly with our end-user to establish what they can afford and how much they will be willing to pay.

That said, it is still informative to thoroughly evaluate a technology to see how it fits into the panoply of options. This is exactly what we are attempting to do with our wind turbine and the testing we intend to conduct at the NASA-Ames research center. Our goal is to establish just how much mechanical power one can reasonably attempt to collect from the wind with a small, family-sized, vertical-axis wind turbine using the most promising blade designs we’ve encountered. Using the data we collect we should be able to estimate just how much it will cost per watt to simply capture the energy before turning it into electricity (this is based on the costs associated with building the turbine: blades, bearing, shaft, etc.). From there we will be able to predict just how much money we can afford to spend on the generator and circuitry and still achieve the price point our partner, AIDG, will help us establish (is is truly $1/watt or a more achievable $10/watt?). The amount of money available for the generator and circuitry based upon this price point will directly inform just how feasible this product may be for rural Guatemalan villages.

It could well be that the raw materials alone for the alternator simply price the whole concept above our acceptable price target; it is, after all, notoriously difficult to scale wind generation down and keep it cost-effective. As a rule of thumb, the larger a turbine is the less the electricity it generates costs per watt. We hope our tests prove that small, vertical-axis turbines are a promising avenue for additional design research (we’re fans of wind energy, after all), but the numbers won’t lie and should allow us to provide AIDG with a realistic evaluation as to how good an investment their continued development appears to be.