Researchers have long known that general aviation (GA) aircraft exhibit some of the most intense and potentially damaging sound environments to a pilot's hearing. Yet, another potentially more ominous result of this noise-intense environment is the masking of the radio communications. Radio communications must remain intelligible, as they are imperative to the safe and efficient functioning of the airspace, especially the airspace surrounding our busiest airports, Class B and Class C. However, the high amplitude, low frequency noise dominating the GA cockpit causes an upward spreading of masking with such inference that it renders radio communications almost totally unintelligible, unless the pilot is wearing a communications headset. Even with a headset, some researchers have stated that the noise and masking effects overcome the headset performance and still threaten the pilot's hearing and overall safety while in the aircraft. In reaction to this situation, this experiment sought to investigate the effects which active noise reduction (ANR) headsets have on the permissible exposure levels (PELs), speech intelligibility, workload, and ultimately the pilot's performance inside the cockpit. Eight instrument-rated pilot participants flew through different flight tasks of varying levels and types of workload embedded in four 3.5 hour flight scenarios while wearing four different headsets. The 3.5 hours were considered long duration due the instrument conditions, severe weather conditions, difficult flight tasks, and the fatiguing effects of a high intensity noise environment. The noise intensity and spectrum in the simulator facility were specifically calibrated to mimic those of a Cessna 172. Speech intelligibility of radio communications was modified using the Speech Transmission Index (STI), while measures of flight performance and workload were collected to examine any relationships between workload, speech intelligibility, performance, and type of headset. It is believed that the low frequency attenuation advantages afforded by the ANR headset decreased the signal-to-noise ratio, thereby increasing speech intelligibility for the pilot. This increase may positively affect workload and flight performance. Estimates of subjective preference and comfort were also collected and analyzed for relevant relationships. The results of the experiment supported the above hypotheses. It was found that headsets which incorporate ANR technology do increase speech intelligibility which has a direct inverse influence on workload. For example, an increase in speech intelligibility is seen with a concomitant decrease in pilot workload across all types and levels of workload. Furthermore, flight task performance results show that the pilot's headset can facilitate safer flight performance. However, the factors that influence performance are more numerous and complex than those that affect speech intelligibility or workload. Factors such as the operational performance of the communications system in the headset, in addition to the ANR technology, were determined to be highly influential factors in pilot performance. This study has concluded that the pilot's headset has received much research and design attention as a noise attenuation device. However, it has been almost completely overlooked as a tool which could be used to facilitate the safety and performance of a general aviation flight. More research should focus on identifying and optimizing the headset components which contribute most to the results demonstrated in this experiment. The pilot's headset is a component of the aviation system which could economically improve the safety of the entire system. / Ph. D.

Control of radiated acoustic noise is vital to the survivability and the detectability of submersible watercraft. Two primary sources of radiated fluid noise in submersible vessels are the boundary layer turbulence along the forebody and propulsor fluid-structure interaction. The propulsor contains several locations of such interaction, one of which was investigated in this research. Specifically, this research focused on experimentally investigating active flow control techniques to reduce rotor-stator interaction noise sources. Two of the three flow control configurations applied to the flow involved the application of active flow control to the leading edge of a single exit guide vane (EGV) mounted downstream of a seven-bladed rotor. The leading edge blowing configuration (LEB) consisted of a single jet expelled from the leading edge of the EGV against the oncoming flow. This interaction between the wake and jet should offset or disrupt the coherency of any incoming flow structures. The second active flow control method applied to the EGV involved a tangential blowing configuration (TB) where two symmetric tangential jets were used to create an insulating fluid layer that reduced the effect of passing flow structures on the EGV. The final flow control design was the implementation of trailing edge wake filling on a three bladed rotor. A rotor was designed to ingest lower velocity flow from the hub and pump the fluid out of a blowing slot at the blade trailing edge. The blowing slot was concentrated on the outer third of the blade span in order to maximize pumping effect. In order to quantify the effects of the active flow control techniques on rotor-stator interaction, the fluctuating lift force on the EGV was measured. Since this fluctuating force serves as a primary acoustic source, the effects of the active flow control on the radiated interaction sound can be estimated. These active flow control techniques were intended for reduction of blade passing frequency tonal sound radiation. The LEB configuration showed minor changes in overall spectral response; however, there was no significant reduction in forcing at the BPF measured. Similarly the TB configuration also yielded no measurable change in BPF tonal forcing. The first generation design of the self-pumping rotor also proved to have problems. Experiments showed that the application of the flow control on the self-pumping rotor did not generate the expected increase in torque demand or changes in the tonal forcing on the EGV. Field alterations to the rotor were unable to improve the performance; therefore, the conclusion became that the initial design was unable to pump fluid due to excessive pressure losses. Further design iterations are required to perfect the functionality of the self-pumping rotor. / Master of Science

Key areas of noise sources are investigated through comparison of eddy convection velocity and turbulence measurements in three-stream nozzles. A Time-Resolved Doppler Global Velocimetry (TR-DGV) Instrument was applied to the Nozzle Acoustic Test Rig (NATR) at NASA's Aero-Acoustic Propulsion Lab (AAPL) to measure convection velocity. Particle image velocimetry (PIV) measurements provided mean velocity and turbulence intensity. Eddy convection velocity results were obtained from the TR-DGV data for three-stream nozzle configurations using a cross-correlation approach. The three-stream cases included an axisymmetric and an asymmetric nozzle configuration. Results of the VT TR-DGV convection velocity were compared to NASA PIV mean and turbulence intensity data. For the axisymmetric case, areas of high convection velocity and turbulence intensity were found to be from 4 to 6 diameters downstream. Comparison of convection velocity between the axisymmetric and offset case show this same region as the greatest reduction in convection velocity due to the offset. These findings suggest this region along the centerline near the end of the potential core is an important area for noise generation with jets and contribute to the noise reductions seen from three stream offset nozzles. An analysis of a one-dimensional wavepacket model was completed to provide understanding of the effect of the various convection velocities seen in the flow. Comparison of a wavepacket with a convection velocity of 0.6Uj to a wavepacket with a convection velocity of 0.8Uj showed that an increase in convection velocity shifts the wavenumber spectrum to higher wavenumbers as expected. It was also observed that for the higher convection velocity wavepacket, higher frequencies are more acoustically efficient, while mid frequencies are the most efficient radiators in the lower convection velocity case. Using mean velocity, turbulence intensity, and convection velocity areas of likely to generate noise are identified and possible fundamental mechanisms responsible for the noise generation are discussed. / Master of Science

Two experiments were conducted to investigate Adaptive Noise Cancelling and Cepstrum echo removal post-processing techniques on acoustic data from a linear microphone array in an anechoic chamber. A point source speaker driven with white noise was used as the primary signal. The first experiment included a background speaker to provide interference noise at three different Signal-to-Noise Ratios to simulate noise propagating down a wind tunnel circuit. The second experiment contained only the primary source and the wedges were removed from the floor to simulate reflections found in a wind tunnel environment. The techniques were applicable to both signal microphone and array analysis. The Adaptive Noise Cancellation proved successful in its task of removing the background noise from the microphone signals at SNRs as low as -20 dB. The recovered signals were then used for array processing. A simulation reflection case was analyzed with the Cepstral technique. Accurate removal of the reflection effects was achieved in recovering both magnitude and phase of the direct signal. Experimental data resulted in Cepstral features that caused errors in phase accuracy. A simple phase correction procedure was proposed for this data, but in general it appears that the Cepstral technique is and would be not well suited for all experimental data. / Master of Science

The research described herein was undertaken to determine how masked thresholds changed when individuals wore an active noise reduction (ANR) hearing protection device (HPD), a passive HPD, or no HPD. An ANR earmuff, a passive earmuff, and a user-molded foam earplug were tested in two types of noises (pink and red) at two different noise levels (85 dBA and 100 dBA). The signal used was an industry-standard backup alarm. The experimental design was completely within-subjects. An ascending method of limits was used to obtain 15-20 correct positive responses, which were then averaged to obtain the masked thresholds for each treatment condition. A visual probability monitoring task was incorporated in the experimental design to provide a loading task for the participants. In addition to masked thresholds, comfort and mental workload were assessed. Finally, participants were asked to rank each of the three HPDs with respect to their perceived ability to facilitate hearing the signal in noise. Results indicated that in 85 dBA noise, masked thresholds were lower when hearing protection devices were worn, compared to the unoccluded condition. Additionally, the results indicated that the ANR device provided a significant advantage (lower masked thresholds) over the passive earmuff in the low-frequency biased red noise (across both noise levels) and the 100 dBA noise level (across both noise spectra). However, the ANR earmuff exhibited no significant advantage over the user-molded foam earplug in any of the conditions. Rather, the user-molded foam earplug produced significantly lower masked thresholds at 100 dBA. The results also indicated that there was no difference between the three devices in their perceived ability to facilitate detection of the signal. There was also not a significant difference in comfort ratings between the three HPDs, although there were several complaints about the comfort of the ANR earmuff during the experiment. / Master of Science

Speech processing in the automotive environment is a challenging problem due to the presence of powerful and unpredictable nonstationary noise. This thesis addresses two detection problems involving both nonstationary noise signals and nonstationary desired signals. Two detectors are developed: one to detect passing vehicle noise in the presence of speech and one to detect speech in the presence of passing vehicle noise. The latter is then measured against a state-of-the-art voice activity detector used in telephony. The process of compiling a library of recordings in the automobile to facilitate this research is also detailed.

No description available.

Scale modeling has been commonly used for architectural acoustics but use in other noise control areas is nominal. Acoustic scale modeling theory is first reviewed and then feasibility for small-scale applications, such as is common in the electronics industry, is investigated. Three application cases are used to examine the viability. In the first example, a scale model is used to determine the insertion loss of a rectangular barrier. In the second example, the transmission loss through parallel tubes drilled through a cylinder is measured and results are compared to a 2.85 times scale model with good agreement. The third example is a rectangular cuboid with a smaller cylindrical well bored into it. A point source is placed above the cuboid. The transfer function was measured between positions on the top of the cylinder and inside of the cylindrical well. Treatments were then applied sequentially including a cylindrical barrier around the well, a membrane cover over the opening, and a layer of sound absorption over the well. Results are compared between the full scale and a 5.7 times scale model and correlation between the two is satisfactory.

Mufflers and silencers are commonly used to attenuate noise sources such as internal combustion engines and HVAC systems. Typically, these environments contain mean flow that can affect the acoustic properties of the muffler components and may produce flow generated noise. To characterize the muffler performance, common metrics such as insertion and transmission loss and noise reduction are used in industry. Though transmission loss without flow is often measured and is a relatively simple bench top experiment and useful for model validation purposes, mean flow can significantly affect the muffler performance. There are a few existing and commercial transmission loss rigs that incorporate flow into the measurement procedure. These rigs are useful for model verification including flow but do not predict how the muffler will perform in the system since the source, termination, and pipe lengths significantly impact performance. In this research, the development of an insertion loss test rig is detailed. This testing strategy has the advantage of being simpler, quantifying the self-generated noise due to flow, and taking into account the effect of tailpipe length and a realistic termination. However, the test does not include the actual source and is not as useful for model validation. An electric blower produces the flow and a silencer quiets the flow. Loudspeakers are positioned just downstream of the flow silencer and they are used as the sound source. The low frequency source is a subwoofer installed in a cylindrical enclosure that includes a conical transition from speaker to pipe. Special care is taken to reduce any flow generated noise. Qualification of the system is detailed by comparing the measured transmission loss, noise reduction, and insertion loss to one-dimensional plane wave models. The results demonstrate that the developed rig should be useful as a muffler evaluation tool after a prototype has been constructed. The rig can also be used for transmission loss and noise reduction determination which will prove beneficial for laboratory testing.

Speech is a fundamental means of human communication. In the last several decades, much effort has been devoted to the efficient transmission and storage of speech signals. With advances in technology making mobile communication ubiquitous, communications anywhere has become a reality. The freedom and flexibility offered by mobile technology brings with it new challenges, one of which is robustness to acoustic background noise. Speech enhancement systems form a vital front-end for mobile telephony in noisy environments such as in cars, cafeterias, subway stations, etc., in hearing aids, and to improve the performance of speech recognition systems. In this thesis, which consists of four research articles, we discuss both single and multi-microphone approaches to speech enhancement. The main contribution of this thesis is a framework to exploit available prior knowledge about both speech and noise. The physiology of speech production places a constraint on the possible shapes of the speech spectral envelope, and this information s captured using codebooks of speech linear predictive (LP) coefficients obtained from a large training database. Similarly, information about commonly occurring noise types is captured using a set of noise codebooks, which can be combined with sound environment classi¯cation to treat different environments differently. In paper A, we introduce maximum-likelihood estimation of the speech and noise LP parameters using the codebooks. The codebooks capture only the spectral shape. The speech and noise gain factors are obtained through a frame-by-frame optimization, providing good performance in practical nonstationary noise environments. The estimated parameters are subsequently used in a Wiener filter. Paper B describes Bayesian minimum mean squared error estimation of the speech and noise LP parameters and functions there-of, while retaining the in- stantaneous gain computation. Both memoryless and memory-based estimators are derived. While papers A and B describe single-channel techniques, paper C describes a multi-channel Bayesian speech enhancement approach, where, in addition to temporal processing, the spatial diversity provided by multiple microphones s also exploited. In paper D, we introduce a multi-channel noise reduction technique motivated by blind source separation (BSS) concepts. In contrast to standard BSS approaches, we use the knowledge that one of the signals is speech and that the other is noise, and exploit their different characteristics. / QC 20100929