Selected Publications
This paper describes noise measurements taken of the new Boeing T-7A Red Hawk trainer aircraft, which uses a single F404 afterburning turbofan engine. The extensive measurement satisfies the American National Standards Institute/Acoustical Society of America standard S12.75-2012 for ground run-up for future environmental impact assessment and includes additional locations around the aircraft to understand exposure by maintenance personnel. A large near-field array was also deployed to shed light on phenomena that are not generally seen in the measurement of laboratory-scale jets, such as the presence of spatiospectral lobes. Initial data analysis shows they are of high fidelity and contain similar phenomena as other recent high-performance jet aircraft noise measurements, including evidence of large and fine-scale noise radiation, broadband shock-associated noise, spatiospectral lobing at multiple engine powers, an upstream shifting of overall level directivity with engine power, and appreciable shock content in the measured waveforms. Further analysis of this dataset will add to the understanding of full-scale, high-speed jet noise and allow comparisons to similar numerical simulations and laboratory-scale measurements.
Crackle is a perceptual feature of supersonic jet noise that is related to the presence of acoustic shocks. This study investigates the apparent source locations of the steepest shocklike events in the noise field of a high-performance military jet aircraft using an event-based time-domain beamforming method. This method uses the cross correlation between adjacent microphones to determine the angle of propagation of an ensemble of shock-related events within the time waveform. This angle of propagation is then traced back toward the source to find the apparent source location. Based on the propagation angle, derivative skewness, and overall sound pressure level, the microphone positions along the array are sorted into six groups. Shock events from groups related to crackle perception in the near field originate anywhere from 2 to 14.5 m downstream along the nozzle lip line, with distributions that shift downstream and broaden with increasing engine power. The crackle-related events appear to be generated by Mach wave radiation and large-scale turbulence structure noise.
This paper describes the development of an automated classification algorithm for detecting instances of focused crowd involvement present in crowd cheering. The purpose of this classification system is for situations where crowds are to be rewarded for not just the loudness of cheering, but for a concentrated effort, such as in Mardi Gras parades to attract bead throws or during critical moments in sports matches. It is therefore essential to separate non-crowd noise, general crowd noise, and focused crowd cheering efforts from one another. The importance of various features—both spectral and low-level audio processing features—are investigated. Data from both parades and sporting events are used for comparison of noise from different venues. This research builds upon previous clustering analyses of crowd noise from collegiate basketball games, using hierarchical clustering as an unsupervised machine learning approach to identify low-level features related to focused crowd involvement. For Mardi Gras crowd data we use a continuous thresholding approach based on these key low-level features as a method of identifying instances where the crowd is particularly active and engaged.
Since the Morfey-Howell Q/S was introduced as a single-point frequency-domain nonlinearity indicator for propagation of intense broadband noise [AIAA J. 19, 986–992 (1981)], there has been debate about its validity, utility, and interpretation. In this Letter, the generalized Burgers equation is recast in terms of specific acoustic impedance along with linear absorption and dispersion coefficients, normalized quadspectral density (Q/S), and newly proposed normalized cospectral density (C/S). The formulation leads to a rather straightforward interpretation in which Q/S and C/S, respectively, represent the additional absorption and dispersion at a locale, produced by the passage of a finite-amplitude wave.
Outdoor ambient acoustical environments may be predicted through machine learning using geospatial features as inputs. However, collecting sufficient training data is an expensive process, particularly when attempting to improve the accuracy of models based on supervised learning methods over large, geospatially diverse regions. Unsupervised machine learning methods, such as K-Means clustering analysis, enable a statistical comparison between the geospatial diversity represented in the current training dataset versus the predictor locations. In this case, 117 geospatial features that represent the contiguous United States have been clustered using K-Means clustering. Results show that most geospatial clusters group themselves according to a relatively small number of prominent geospatial features. It is shown that the available acoustic training dataset has a relatively low geospatial diversity because most training data sites reside in a few clusters. This analysis informs the selection of new site locations for data collection that improve the statistical similarity of the training and input datasets.
Flushing a vacuum-assisted toilet generates noise levels that can be disturbing both to users and those nearby. Peak radiated noise levels correlate with the time when the valve opens and closes, while the noise levels when the valve is com- pletely open are also relatively high. Significant noise ranges between 300 Hz and 3000 Hz. It was hypothesized that increasing the in-tube distance between the flush valve and the bowl in addition to increasing the bend radius of the tube would reduce radiated noise levels. These modifications resulted in a reduction of about 14 dB in the radiated noise during the valve opening and closing in ad- dition to a reduction of about 5 dB while the valve is completely opened. Inter- mediate results of varying the tube length and bend radius are presented to show their effects on the radiated sound levels. Two tube inserts were designed to fit (1) underneath and (2) behind the toilet in a compact manner. They were tested to show that they maintain noise control performance without modifying any other part of the toilet.