Measurement and Prediction of Rotor Noise Sources for sUAS in Outdoor and Laboratory Environments

This work provides an experimental analysis of the acoustic footprint of a hexacopter in hover and low speed forward flight, comparison of aerodynamic performance and noise of eVTOL rotors operating in multiple facilities, and analysis of the noise associated with an outrunner brushless DC motor. Empirical and low-order models are used to predict aerodynamic performance, tonal noise, and broadband noise for isolated eVTOL rotors. In addition, a low noise, swept rotor design was evaluated. The acoustic footprint of a DJI Matrice 600 Pro hexacopter in hover and low speed forward flight was measured in the Virginia Tech Drone Park. The noise radiated by this vehicle was found to be dominated by tonal noise at low frequencies and dominated by broadband noise at high frequencies indicating that both are important when assessing the noise of these aircraft. Three distinct regions were observed in the frequency spectra of the noise. A-weighting measured acoustic spectra highlighted the importance of the mid-frequency broadband noise, in particular. The radiated noise in hover was also found to be similar to the noise of the vehicle during low-speed flyovers. Given this, significantly less complex measurements of an aircraft in hover or those associated with a rotor at static conditions may be used to assess the footprint of an eVTOL aircraft in low speed forward flight. The total vehicle noise was then decomposed by studying the performance and noise of isolated eVTOL rotors in multiple facilities and under different operating conditions. Facility effects on performance and noise were first assessed by experimentally studying two commercially available eVTOL rotors in an enclosed anechoic environment and an open environment. For experimental measurements that were conducted in the anechoic chamber, recirculation effects were shown to increase harmonic amplitudes more than 8 dB. Varying solidity screens were placed in the downstream wake of each rotor to delay the onset of recirculation. Placing the screens in the wake did not produce a noticeable effect on or delay recirculation within the confined testing environment. Measurements of the BPF and higher order harmonics of each rotor were found to be much more consistent in time when testing outdoors in an open-air environment. Amplitudes of these tones were also found to be like that of the spectral levels of the measurements conducted in the anechoic chamber once recirculation had been established. While the tonal levels were much more consistent throughout each measurement in the open-air environment, a significant amount of background noise was present and made characterizing the noise at low frequencies difficult. Environmental factors, mainly windspeed, were also found to impact the noise measurements which also added difficulty in characterizing the noise of the two tested rotors. In indoor facilities, the rotor inflow becomes contaminated due to recirculation shortly after the rotor reaches steady state and spectral levels of tones increased with increasing spectral averaging times. In outdoor environments, the inflow to the rotor disc becomes distorted due to changing wind conditions and turbulence in the atmosphere. Spectral levels of tones in the outdoor environment remained consistent in amplitude but exceeded those of the anechoic chamber significantly. Given this, environmental factors and recirculation were found to both increase the higher order harmonics. To mitigate these facility effects, measurements of force and noise were also conducted for the same two rotors in an anechoic open jet. Additionally, measurements were also conducted for a commercially available rotor along with a newly designed low noise swept rotor. Each of these rotors were tested in the anechoic open jet facility at static conditions and with the tunnel on. These measurements were accompanied with predictions of aerodynamic performance and tonal and broadband self-noise. BEMT was used to predict aerodynamic performance. Tonal noise associated with the rotor blade loading and thickness was predicted using F1A and rotor broadband self-noise was predicted using the model of BPM. The measured noise in this facility along with that from measurements in the anechoic chamber and outdoor environment were separated into tonal and broadband components by applying a phase averaging technique to the measured acoustic pressure time history. These results also show that in the indoor facility that the noise produced at the BPF is dominated by tonal sources, but the higher order harmonics can be attributed to broadband interactions particularly at static conditions. Broadband noise was drastically reduced by driving the tunnel at minimal inflow for the smallest rotor studied (R_tip= 120 mm). For the larger rotors (R_tip≥ 267 mm) broadband noise associated with BWI or TIN were not mitigated at low inflow speeds. Predictions of tonal noise at the BPF were within 3 dB for all observer locations when considering the smallest rotor studied. Predictions of the measured directivity at the BPF for the larger rotors were inaccurate although predictions of thrust agreed with the measured. The largest rotors tested were equal in diameter to that of the open jet inlet. Thus, the limits of the testing facility were exceeded and increased noise was produced as the rotor blades interacted with the shear layer of the open jet. Directivity patterns of each rotor were also found to vary with increasing rotational rate. Overall, these results show that for analyzing the noise at hover conditions, introducing a small amount of inflow may be a good option when trying to understand the tonal noise and allows one to characterize the tonal noise independent of the broadband. However, this was also shown to be heavily dependent on the rotor diameter with regards to the open jet inlet and experimentalist must take this into consideration. While these measurements provide an analysis of the noise in hover and low speed ascent, they do not assess the noise of the vehicle operating in forward flight. In forward flight the rotors are subjected to edgewise flows which have an effect on the radiated noise thus analyzing the noise of these rotors operating at an angle of attack to the incoming flow was assessed. These effects were investigated by experimentally measuring the performance and noise of the smallest rotor studied when operating at a yaw relative to the incoming flow. For increasing yaw at the examined wind tunnel velocities, the measured thrust was found to converge to the value for zero inflow. Contours of SPL as a function of yaw angle for no inflow and an inflow speed of 8 m/s showed spectral levels to be minimal for an in-plane observer from 5×BPF to 30×BPF. The broadband noise was found to increase significantly for increasing yaw angle and tunnel inflow speed. These results show once again that the broadband noise is especially important during forward flight and new methods that consider wake interaction are needed to predict the noise in this flight regime. The rotor geometric parameter of sweep was also assessed from measurements in the anechoic open jet by comparing the aerodynamic performance and noise of a commercially available 762 mm diameter CF30x10.5 T-motor eVTOL rotor to that of an in house designed low noise swept rotor. The addition of sweep was found to reduce noise associated with BWI or TIN as the separated broadband noise was found to be less than that of the commercially available rotor. Comparison of thrust at static conditions and with increasing advance ratios showed both rotors to have similar performance, thus the addition of sweep was effective at reducing noise without sacrificing performance. Lastly, the noise associated with the electric drive system of these aircraft which consists of an ESC and brushless DC motor was analyzed. Acoustic measurements were made with and without an acoustic enclosure installed on a brushless DC motor and was found to be effective at reducing noise associated with the electric motor. The effects of two ESC's as well as their switching rates were also studied. The noise was found to be similar for both ESCs at low frequencies. At high frequencies the measured noise spectrum was found to be different when controlling the motor with different ESC's and a higher switching rate was found to reduce the noise with increasing switching rate although not completely monotonically.

Doctor of Philosophy

A new class of multi-rotor VTOL electric aircraft is becoming a dominant advanced vehicle concept. Urban Air Mobility (UAM) vehicles are designed for short routes within urban environments carrying only a few passengers during each flight. Other smaller Unmanned Aerial Systems (UAS) are increasingly being used for delivery services or to perform tasks which are more easily accessed with this technology like inspection or photography. Thus, these vehicles are expected to operate in close proximity to the general populace exposing it to aircraft noise which is currently limited to communities surrounding airports. For successful integration into the airspace with minimal community annoyance, the mechanisms responsible for generation of the noise must be understood. Traditionally, for conventional rotorcraft (one main rotor), the tonal noise has been more of a concern than the broadband component. eVTOL vehicles are often equipped with multiple rotors that are lightly loaded and operate at lower tip speeds which can be time varying. Thus, there is an increased significance of broadband noise. Lastly, these aircraft are equipped with an electric drive system that gives rise to an additional noise source that is not present for conventional aircraft. Best practices for measuring eVTOL noise are not currently established. Measurement of eVTOL rotor noise is complicated by the increased significance of the broadband sources. These have been shown to be facility dependent. Given this, there is a need for high quality experimental data and an analysis of experimental data in multiple facilities for these rotors and drive systems. Capabilities of traditional models to predict conventional rotorcraft noise also need to be assessed for these rotors. These two issues have been assessed in this work by first assessing the character of an eVTOL aircraft in hover and low speed flyovers. Both tonal and broadband components of the radiated noise were found to be significant. A-weighting, which is a metric used to assess the response of the human ear to the radiated noise showed increased significance of the broadband noise. This was followed by a characterization of the noise of isolated eVTOL rotors in multiple environments. Facility effects were addressed, and a low order prediction model was developed using methods that are traditionally used to predict noise associated with conventional rotorcraft. Lastly, the noise associated with the electric drive system of these vehicles was assessed and recommendations on how to reduce this source of noise were made. These results can be used to guide experimentalists when performing measurements of eVTOL rotor noise at static conditions and provide an eVTOL rotor noise data set that can be used to validate existing and forthcoming aerodynamic and acoustic prediction methods.