UC Davis

The research objective of this dissertation is to advance the understanding of rotorcraft broadband noise and reduce the noise impacts. Rotorcraft broadband noise has recently become a critical topic for vertical take-off and landing (VTOL) aircraft, due to the rapid progresses in Advanced Air Mobility (AAM) technologies. However, the current noise assessment tools of VTOL broadband noise rely heavily on semi-empirical models, which are short of both prediction accuracy and flow physics. To close this gap, a state-of-the-art prediction tool, namely UCD-QuietFly, is developed to assess rotor broadband noise using physics-based approaches. Extensive validations of UCD-QuietFly are performed against experiments. The effects of rotor design parameters on broadband noise are studied. Broadband noise impacts are investigated on Urban Air Mobility (UAM) conceptual designs, quiet helicopter designs, and small-scale drones. Finally, the noise reduction technique using trailing-edge serrations is analyzed.

The first part of this dissertation investigates the effects of rotorcraft design and operating parameters on trailing-edge noise. A rotor trailing-edge noise prediction method is first developed. It is found that helicopter broadband noise scales with the 4.5th to 5.0th power of the tip Mach number in which the range is determined by the typical helicopter collective pitch angle in operation. Detailed trend analyses of noise levels as a function of frequency are presented in terms of the collective pitch angle, twist angle, rotor solidity, rotor radius, disk loading, and number of blades.

Second, broadband noise of multi-rotor UAM vertical take-off and landing (VTOL) vehicles is studied. The multi-rotor broadband noise prediction capability is developed. It is found that UAM VTOL vehicles' broadband noise is important in the high-frequency range. For the same mission specifications, broadband noise is found to be higher for VTOL designs with more rotors. Multi-rotor vehicles at the same rotational speeds have weaker amplitude modulations than a single rotor, which demonstrates the benefits of using multiple rotors in terms of noise annoyance.

Third, tonal and broadband noise are studied for rotor designs used on UAM vehicles. The rotor aerodynamics in edge-wise forward flight is calculated by an in-house code. With the forward flight capability developed in UCD-QuietFly, it is found that broadband noise is the dominant noise source for the rotor designs with low tip speeds and fewer blades, while tonal noise is dominant for the high-tip-speed designs. A low tip speed and more blades are found to be the preferable design features in terms of psychoacoustic metrics.

Fourth, a physics-based broadband noise prediction approach is applied for small-scale drone rotors. LBL-VS noise is found to be an important noise source for untripped blades. The effect of leading-edge back-scattering in broadband noise is found to be important for small-scale rotors at low frequencies. Finally, at the same thrust, the ideally twisted rotor generates slightly higher broadband noise than the linearly twisted rotor, while the tapered blade tip is shown to reduce the broadband noise levels significantly.

Fifth, machine learning-based fast-predicting models of rotorcraft trailing-edge broadband noise are developed, using artificial neural network (ANN) and linear regression. It is found that the ANN model accurately captures the variations of the noise levels, and the linear regression models are also capable of predicting the general trends of noise levels.