Theses and Dissertations at Montana State University (MSU)
Permanent URI for this communityhttps://scholarworks.montana.edu/handle/1/732
Browse
5 results
Search Results
Item Assessing nonlinearity and memory extent in audio systems(Montana State University - Bozeman, College of Engineering, 2021) Hoerr, Ethan Randall; Chairperson, Graduate Committee: Robert C. MaherCreating digital models of existing audio devices is useful for increasing access to audio effects and for preserving audio history. In the work covered by this dissertation, we investigate the use of Volterra series modeling to assess the degree of nonlinearity of a system and time-delayed mutual information (TDMI) to estimate the length of the recovered impulse response. Using an arctangent function as an example system, comparing empirically generated Volterra series models containing anywhere from first- to fourth-order system kernels revealed that including the odd-ordered first and third kernels yielded the best-performing model. We propose that this benchmarking method can aid a system modeler by elucidating details about a system's nonlinear behavior. We also assess the utility of time-delayed mutual information (TDMI) as a method for revealing which samples of a recovered impulse response of a nonlinear system are significant. As a nonlinear metric of correlation between an input signal x[n] and output signal y[n], the TDMI method in MATLAB simulations accurately predicted the significant samples of delay lines, FIR moving average filters, and Schroeder all-pass filters. The TDMI method was less informative when applied to IIR low pass filters with and without an arctangent function appended to the output. Finally, we applied the TDMI approach to a real-world audio device, a distortion effects pedal designed for electric guitar players. In the presence of increasing nonlinear distortion, the calculated TDMI curve took the shape of a pronounced peak starting at T = 0 samples delay between x[n] and y[n], with T increasing as the distortion increased. A similar phenomenon was observed when lowering the pedal's low-pass filter cutoff frequency from 36.7 kHz to 620 Hz; in the 620 Hz test, the TDMI peak was significantly lower than the other test cases and featured a more gradual decay to the estimator bias noise floor. In summary, we demonstrated that Volterra series models are useful for assessing the degree of nonlinearity of a system and that time-delayed mutual information can inform which samples of a recovered impulse are significant. Both of these insights can aid in deciding how many Volterra series kernels and how much kernel memory to include when creating a black box system model.Item An acoustic emission and hygrothermal aging study of fiber reinforced polymer composites(Montana State University - Bozeman, College of Engineering, 2019) Newhouse, Kai Jeffrey; Chairperson, Graduate Committee: David A. MillerFiber reinforced polymer matrix composites are a premier choice for offshore wind turbines and Marine Hydro-Kinetic Devices, which operate in harsh and isolated marine environments. These factors combined with decades long target service life make imperative the understanding of damage mechanisms and the environmental effects thereof. Acoustic emission monitoring is a research technology that uses specialized sensors to detect transient elastic waves in a material which originate from damage sources. Waveform parameters have been correlated with different damage mechanisms in fibrous composites. A diverse set of fiber-matrix combinations configured into a variety of layups totaling more than 30 laminates were mechanically tested in quasi-static uniaxial tension while monitoring acoustic emission. A subset of these materials was aged prior to testing in an artificial marine environment by soaking in a water bath of simulated seawater at 50 degrees Celsius. Various acoustic emission waveform parameters were investigated with respect to expected damage between layups and possible material-based differences. Among the conditioned material set, mechanical changes from moisture absorption shows mixed levels of degradation among different material systems. Moduli were generally unaffected with a few minor decreases. Strengths were reduced by as much as 41%, and failure strains fell as much as 47%. From acoustic emission investigation, good correlation is found between Fast Fourier Transform peak spectral frequency bands and expected damage mechanisms between layups. Material based peak frequency differences are found exclusively in interphase failures (de-bond and fiber pullout). Layup-based correlations in conjunction with elastic wave theory were used to put forth new frequency band ranges associated with damage types.Item Combining acoustic emission and guided ultrasonic waves for global property prediction and structural health monitoring of glass fiber composites(Montana State University - Bozeman, College of Engineering, 2018) Murdy, Paul; Chairperson, Graduate Committee: David A. MillerSince the turn of the century, wind turbines have been rapidly growing in size and are projected to continue growing as the technology develops. These increases in size have led to increased failure rates of the glass fiber composite turbine blades. Because of this, it is of utmost importance to understand failure mechanisms in glass fiber composites and investigate new approaches to predicting failures. This has led to advancements in structural health monitoring of large composites structures by applying sophisticated sensing technologies, in attempts to evaluate material damage states and predict structural failures before they occur. This research has taken a novel approach to apply multiple ultrasonic monitoring techniques, in the form of acoustic emission and guided ultrasonic waves, simultaneously to the mechanical testing of glass fiber reinforced composite laminates. Testing of the composite laminates was conducted in the form of increasing load-unload-reload static tension tests and tension-tension fatigue tests, to measure modulus degradation of the laminates while applying the monitoring techniques. Acoustic emission was used to detect damage events that occurred within laminates in real-time and guided ultrasonic waves were applied periodically to the laminates to observe changes in wave propagation and relate back to damage severity within the laminates. Furthermore, the acoustic emission and guided ultrasonic wave datasets were combined and used to train multivariate regression models to predict modulus degradation of the laminates tested, with no prior knowledge of the laminates' loading histories. Overall, the predictive models were able to make good predictions and showed the potential for combining multiple monitoring techniques into singular systems and statistical predictive models. This research has shown that the combination of the two measurement techniques can be implemented for more accurate and reliable monitoring of large composite structures than the techniques used individually, with minimal additional hardware. Ultimately, this research has paved the way for a new form of smart structural health monitoring, with superior predictive capabilities, which will benefit the renewable energy through reducing maintenance and repair costs and mitigating the risk of wind turbine blade failures.Item Observing variation of acoustical characteristics of several common firearms in a quasi anechoic environment at a high sampling rate(Montana State University - Bozeman, College of Engineering, 2016) Routh, Tushar Kanti; Chairperson, Graduate Committee: Robert C. MaherAudio recordings from a shooting incident may provide crucial information for a criminal investigation. A typical gunshot signal includes two high amplitude and short duration impulsive signature sounds, the muzzle blast, observed in all the gunshot waveforms, and the bullet s shock wave, which can only be detected if the bullet travels at supersonic speed. Acoustic gunshot analysis generally focuses on the study of muzzle blast signals, which last only a few milliseconds. Ideally, gunshot signals needed to be record at a very high sampling rate to reveal the muzzle blast details. Real life gunshot recordings are record with equipment not designed for these high-amplitude sounds. Moreover, the recordings contain the direct sound of the gun along with multiple overlapping signals due to sound reflections from the ground, nearby surfaces, and other obstacles. The resulting reverberant recording may be difficult to interpret. To study the details of these signals in a scientific manner, we have developed a quasi-anechoic procedure to capture gunshot signals at a very high sampling rate (500 kHz samples per second) using 12 microphones covering 180° in azimuth. The recordings are made in an open air environment with a raised shooting platform and microphone position, resulting in sufficient delay between the arrival of the direct sound at the microphones and the arrival of the first reflection (from the ground). The firearms used in this experiment include a Remington 870 shotgun, 308 Winchester rifle, AR15 rifle, and a 22LR rifle. Handguns tested include a Colt 1911A1, Glock 19 with 9mm ammunition, Glock 23, Sig 239, and a Ruger SP101 with both 357 Magnum and 38 Special ammunition. A number of successive shots were record for each of the firearm type. Based on analysis of the recorded data, we find that acoustic gunshot signals vary from one firearm to another in terms of peak sound pressure and Muzzle blast duration. For a given firearm, we observe significant differences in sound level and also Muzzle blast duration as a function of azimuth and find that there is measurable variation in signal details among successive shots from the same firearm.Item Virtual audio localization with simulated early reflections and generalized head-related transfer functions(Montana State University - Bozeman, College of Engineering, 2009) Reed, Darrin Kiyoshi; Chairperson, Graduate Committee: Robert C. MaherIn a natural sonic environment a listener is accustomed to hearing reflections and reverberation. It is conceived that early reflections could reduce front-back confusion in synthetic 3-D audio. This thesis describes experiments which seek to determine whether or not simulated reflections can reduce front-back confusions for audio presented with non-individualized head-related transfer functions (HRTFs) via headphones. To measure the contribution of the reflections, 13 human subjects participated in localization experiments which compared their localization ability with anechoic HRTF processing versus HRTF processing with a single early-reflection. The results were highly subject dependent; some showed improvement while others seemed to be inhibited by the reflections. Statistical analysis of the overall results concluded that a single reflection does not provide a significant difference in localization ability. Although this data rejects the hypothesis of this investigation, some suspicion regarding the contribution of lateral reflections in an auditory environment remains.