MKS INSTRUMENTS MECHANICAL/ELECTRICAL ENGINEERING INTERNSHIP.
STRUCTURAL WIND TURBINE GRADUATE MECHANICAL ENGINEERING RESEARCH
INFORMATION
Client
Boston University Department of Engineering Research
Location
Boston, MA
Date
5/1/25 – 9/2/25
Programs Used
Arduino IDE
CSS
HTML
SolidWorks
SolidWorks Flow CFD
​
Coastal wind turbines experience wind forces greater than that of inland wind turbines. These regions tend to produce stronger and more consistent winds the farther out from shore a turbine is located. Thus, the energy generation from using offshore wind turbines proves to be of greater potential for development and expansion. However, exposure to these intense wind forces produces mechanical strain on the turbine blades causing fatigue in the material. Methods such as prebending help to ensure tower clearance under high load, but are less effective in low velocity wind conditions due to the less efficient aerodynamics.
​
This project seeks to study the effects of gyroscopic precession to provide a counteracting force that resists the bending of the wind turbine blades.
Blade Design and Load Cell
To accurately measure the effects of gyroscopic precession in the environment of a wind turbine, a directional load cell located at the base was chosen with an aluminum reinforced turbine blade. This method was chosen above using a camera and grid ruler to record the deflection of the blade or a resistive surface flex sensor which would utilize too much space or be dependent on the material properties under load. Thus, an Adafruit 1kg load cell strain gauge was chosen as its resolution would be able to accurately detect the smallest of deflections at 24 bits. As a result, the entire length of the blade needed to be rigid to hold the weight of the gyroscope and motor using an aluminum rod and 3D printed PTFE aero. This would ensure any forces leading to blade deformation would be detected at the key stress point at the base. The load cell was also positioned to deflect in the direction of the wind to prevent external forces such as gravity from affecting results.
Gyroscope and Motors
(Turbine Generator)
While mounting the motor with the gyroscope at the tip of the blade encourages more strain on the reinforced blade, the gain in simpler packaging and less complexity of a shaft at this scale was more beneficial. The 3D printed aero profile (modeled after functional wind turbines) includes slots for the 3-phase cables from the brushless motor to be routed. The motor used is a Turnigy D2830-11 Out Runner Brushless Motor typically used for drone applications. Due to their maximum RPM (~19,500 RPM), this made for an ideal driver of the gyroscope as a faster spinning gyro rotor results in a stronger gyroscopic effect. For safety reasons, the target RPM was limited to around 10,000 RPM (~1,047 rad/s as seen in the below equation).
A gyroscope made from a precision aluminium stock with a 3” diameter and ½” thickness would be positioned at the tip of a 23” turbine blade with a 1” pocket to accommodate a high rpm DC brushless motor. The final machined weight was ~0.134 kg which resulted in a moment of inertia (of a disk) of 0.00038 kg • m^2. We can then calculate the estimated precessional angular velocity of the gyroscope to figure out how fast the wind generator must spin. Using the equation below, the following results were obtained.
Programming and BLE Control
Since the gyroscope rotors are themselves rotating at the ends of a wind turbine blade and require control, having to send data and power from the rotating structure proved difficult. Thus, using a wireless communication protocol such as Bluetooth would eliminate the need for unnecessary cable winding or complex contact pads. This would push the choice of a microcontroller to the ESP32 which provides the wireless protocol needed on the board itself. The controller is provided from any computer through an HTML script that connects to a specific ESP32 through a unique UUID address. As such, commands can be executed as well as sensor data being relayed back to the computer.
Data Logging
For data logging, a micro SD card was attached to the central rotating hub via an access slot. Since Bluetooth is low powered and doesn’t have a high bandwidth data rate, data logging directly to a computer would prove difficult with small time intervals. Thus prioritizing the wireless capability to controls would simplify the project and wouldn’t require either a more complex microcontroller or a WIFI protocol requiring a router. However, since the logging process is unobservable, the program is able to update both load cell values as well as the gyroscope RPM values but over long intervals.