Abstract:
A quadrotor was modified by adding wings to the frame to directly compare the flight dynamics characteristics as well as the stability and control derivatives of the quadrotor and its biplane tailsitter variant. The on axis response of the quadrotor and a biplane tailsitter variant were measured through flight test and frequency domain system identification was used for non-parametric and parametric model identification. Identification of the full vehicle dynamics demonstrated that also identifying the motor torque and back-EMF constants from no-load measurements and the remaining motor parameters from a rotor-motor test stand provided the most accurate identified full vehicle model. The motor dynamics were shown to add a pole to the thrust-based responses (roll, pitch, and heave), while the torque based response (yaw) included a pole and a zero. This approach was then used to identify and compare the quadrotor dynamics, tailsitter dynamics, and the total impact of canting the motors. It was found that the presence of the wing added pitch damping to the dynamics and pitch stability became negative. The yaw axis saw an increase in yaw damping derivative, and a reduction in the yaw control derivative to the point where it became difficult to control the aircraft. By introducing cant, both the quadrotor and tailsitter saw large increases in the yaw control derivative. Further, the rotor thrust based moment generation due to cant resulted in the yaw response zero being canceled by the motor dynamics, resulting in a purely first order yaw response. Neither the wing nor cant produced any change in the lateral and heave axes.