What we built for the 2026 Collegiate Wind Competition, how it works, and what went wrong along the way.
A 45 cm, three-blade turbine on a welded steel tower, designed and built in a single semester. It placed 6th of 10 at the 2026 Collegiate Wind Competition.
We did not start from nothing. We read the design reports from the 2025 winners (Cal Maritime, Cal Poly, and CU Boulder), borrowed the ideas that fit a first-year team, and kept everything as simple and as easy to take apart as we could.
Three blades, each 18 cm long and 3D-printed in ABS plastic. Their cross-section is a shape called the A18 airfoil. We compared several shapes in simulation and picked it because it makes a lot of lift for very little drag.
A small gearmotor inside the hub twists all three blades at once through a set of bevel gears. Changing the blade angle changes how much of the wind the blades catch.
A brushless motor run backwards as a generator, mounted right behind the hub. The rotor bolts straight to it with no gearbox in between, and it puts out about 1 volt for every 150 RPM.
A carbon steel pipe TIG-welded to a steel base disc that bolts to the wind tunnel floor. The generator rides on an aluminum rail clamped to the top, so we can slide it to balance the turbine. We sized the tower so it naturally vibrates at 61.2 Hz, just above the 58.3 turns per second the rotor makes at top speed. That way the rotor never reaches the speed that would set the tower shaking.
The power coming out of the generator rises and falls with every gust. A rectifier turns it into DC. Then a maximum power point tracker (MPPT) keeps adjusting how hard the turbine is loaded, so it pulls as much power out of the wind as it can. Ours is a solar MPPT that EasySun donated. The power ends up as heat in a large resistor, which stands in for whatever the turbine would be powering.
An Arduino Uno watches the turbine's voltage and current and decides when to change the blade angle or hit the brake.
The brake works by connecting all three generator wires together. That makes the generator fight its own rotation and stops the rotor. It is wired to fail safe: the Arduino has to keep sending a signal to hold the brake off, so if the controller loses power or a wire breaks, the turbine stops by itself.
Our first printed blades would not have survived the 13 m/s test. We switched to ABS, which handles impacts and bending better.
We started with a helicopter-style swashplate. It was not sturdy enough and was hard to adapt to three blades, so we replaced it with bevel gears that fit inside the 4 cm hub.
Our first brake shorted a single generator phase. That loaded the generator unevenly and caused violent vibration. Shorting all three phases at once brakes smoothly.
As soon as the rotor spun up, the turbine vibrated far more than we expected, so we hit the kill switch instead of risking the hub. The cause was 3D-printed parts that did not line up, from small mistakes in our CAD. Because of that we never measured the full power curve or tested the brake in high wind.
The 2027 turbine is being designed now. Its specs go up here once they are settled, and the design report gets posted after the season.
We take new members at the start of each semester. See how to join.
A group of NC State students forms a team to enter the Collegiate Wind Competition.
Mechanical, Electrical, and Project Development take shape and recruit their first members.
The turbine goes from CAD to a finished machine in one semester, while Project Development writes the wind farm proposal.
Windpack competes at the Collegiate Wind Competition for the first time and finishes 6th of 10.
The 2027 turbine is in design, starting from what the first one taught us.
We take new members at the start of each semester.