Development of a Source Separation Process for Multirotor Aeroacoustic Analysis
Open Access
- Author:
- Rachaprolu, Joel Sundar
- Graduate Program:
- Aerospace Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- November 04, 2022
- Committee Members:
- Amy Pritchett, Program Head/Chair
Eric Greenwood, Thesis Advisor/Co-Advisor
Eric Norman Johnson, Committee Member
Simon W Miller, Committee Member - Keywords:
- Aeroacoustics
Vold-Kalman Filter
De-Dopplerization
Acoustic Hemisphere
UAS and UAM Noise
Noise Source Separation - Abstract:
- With increasing interest in electric Vertical Takeoff and Landing (eVTOL) aircraft for Urban Air Mobility (UAM), many companies are making strides in the design, manufacturing, certification, and operation of these aircraft. Their use depends heavily on community perception and acceptance of these aircraft, with noise generation limited to acceptable levels. These aircraft use Distributed Electric Propulsion (DEP) with multiple rotors to provide the required thrust and power. Thus, their resultant aeroacoustic footprint is more complex than with helicopters. To understand the noise-generating mechanisms of these aircraft: this thesis develops a Source Separation Process (SSP) to separate the noise component produced by each rotor from ground-based flyover acoustic measurements. The SSP is a two-step process that combines time-domain De-Dopplerization with the Vold-Kalman (V-K) order tracking filter. This process can extract rotor components, even when the noise sources continuously change with time, including impulsive noise such as that caused by Blade-Vortex Interaction (BVI). Another advantage of this approach over traditional methods, such as harmonic averaging, is that it preserves the phase and amplitude relationship of acoustic signals throughout the extraction process. The SSP was applied to multiple data sets to validate and highlight the capabilities and limitations. First, the V-K filter of the SSP is applied to wind tunnel data to understand the noise of the top and bottom rotors in coaxial configuration while operating at nearly identical RPM. Second, the de-Dopplerization of the SSP is applied to acoustic flyover measurements of an sUAS octocopter to understand multirotor noise radiation patterns. Third, the complete SSP was applied to acoustic flight test of Bell 430 helicopter to understand the separated noise source variation with time in maneuvering flight. For the separated coaxial rotor data, the bottom rotor has higher levels of loading noise radiating out of the plane due to its operation being in the wake of the top rotor. For the de-Dopplerized octocopter data, the broadband noise is dominant underneath the aircraft whereas the tonal noise is dominant toward the in-plane of the aircraft and the noise is most variable in-plane near the horizon of the aircraft. For the de-Dopplerized and separated flight test data, each separated component has different pulse shapes and directivity trends, the source separation for multiple operating conditions was consistent with the aeroacoustic theory.