Resonator System Property Measurements in the Presence of Noise

In many fields, small volume measurements are critical to validating manufacturing component quality. Small volumes, however, are typically difficult to measure when system geometry is complex. Acoustic methods have come into and out of favor over the last sixty years, and have found interesting applications, but are often accuracy limited due to noise and other measurement uncertainties. Spectral and temporal noise often hinder precision measurements as noise directly corrupts signals of interest. Filtering is useful for limited types of noise, but is ineffective for broad spectrum frequency noise.;To create more accurate non-destructive small volume measurement systems, extracting mechanical system properties from measurement signals is desirable. There are methods for extracting system properties described in literature, including system identification methods, although these methods are more often tailored to black-box control systems where result accuracy and precision are not as critical. With a basic understanding of system dynamics, methods can be tailored to a specific system such that broad-purpose applications are limited, while system-specific applications and accuracy are enhanced.;This research presents a study of acoustic signal composition in the presence of spectral and temporal noise to better understand signal structure and core signal constituents, a method for quantifying magnitudes of amplitude and frequency noise present within a signal, and a flexible parametric numerical model which is able to replicate experimentally acquired signals. This research also describes an improved method, based on principles from both Welch's and Bartlett's methods, for accurately extracting peak resonant frequency from a signal. Additionally, this research presents a novel method for measuring acoustic resonator system volume and effective acoustic neck length using a series of acoustic signals, volume modification devices, and numerical methods.