Cuschieri, Joseph M.

Turbulent pressure fluctuations and acoustical shock
waves formed at pipe discontinuities are the primary source
of flow noise. fhe pipe response is excited by the
fluctuating forces associated with the turbulent pressure
fluctuations. The forcing functions can be determined from
the frequency-wavenumber spectrum of the pressure
fluctuations. A procedure is developed here to obtain the
frequency-wavenumber spectrum due to fully developed
turbulent flow. The data analysis procedures developed in
this study to analyze the pressure fluctuations provide a
good means to determine the frequency-wavenumber spectrum
and represent this data in a clear form. Frequency-wavenumber
spectra have been obtained for simulated pressure
data. In the experimental system designed to collect
turbulent pressure data, it was determined that a recessed
transducer configuration cannot be used in water pipe flow
turbulent pressure fluctuation studies because of the
enhanced turbulence created by the upstream holes.
Therefore, flush mounted transducers are required.

The energy flow and the acoustic radiation of fluid-loaded
panels are investigated using the Energy Accountancy
Concept. The various energy components of the systems are
defined and studied. Each component is a function of the
excitation, the structure, the medium and their coupling.
An energy balance equation is written for the system. This
method is applied to study the acoustic radiation from a
point-excited clamped plate placed on the free surface of a
water tank. The radiation efficiency of the plate is
measured and compared to previous works. The energy
balance equation gives very good results at frequencies
between 50 Hz and 12 kHz. An undefined source of energy
dissipation is observed in one experiment. The results of
this study have shown that the Energy Accountancy Concept
can be used to describe the energy flow in a vibrating
structure under water-loading.

The Energy Accountancy method is used to describe the
response of a system by accounting for the various energy
components in a system, that is components describing the
input energy, the energy dissipated, and the energy
transfered by the system. These components are functions of
quantities that can be determined either through measurement
or finite element analysis of the system. This concept is
used in this study to determine the response of a small
diameter pipe containing two different fluids, air and
water. The results of this study have shown that the Snergy
Accountancy method can be used to describe the response of a
thin walled shell structure with good results. It has also
been shown in this study that in small diameter pipes the
fluid contained by the system can be considered to act as a
reactive medium in the response of the structure.

In shallow water or fluid half-space, the acoustic scattering from a target is significantly different from that of an unbounded medium, due to the multiple reflections occurring between the target and the boundaries. The purpose of this thesis is to investigate the influence of the boundaries on the acoustic scattering of a rigid sphere by means of a superposition method. A minimum number of point sources necessary to accurately model the scattered field is determined in the case of a free medium, a fluid half-space and a waveguide. The free field symmetry vanishes due to the presence of boundaries and, at particular frequencies or scatterer depths, a significant change in the magnitude and spatial distribution of the scattered field occur. In an unbounded medium or fluid half space, the superposition method is shown to give similar results to analytical formulations found in the literature, provided enough point sources are used.

The aim of this thesis is to develop a simulation tool, The 3-D Forward-Look Sonar Simulation Model (3-D-FLSSM), for the 3-D Forward Look Sonar or equivalent that provides insight to the defining characteristics of the sonar system that affect the image quality and the data processing. The simulator includes a representation of the acoustic environment, which incorporates a flat seafloor and spherical target, both of which are assumed to a have small-scale roughness (much less than the acoustic wavelength) associated with them. The backscatter from the target and the seafloor are calculated using the Rayleigh-Rice approximation implementing Kuo's backscattering cross section. The simulator is capable of modeling targets of various shapes and sizes. The 3-D-FLSSM assumes a plane wave approximation and a constant sound velocity throughout the water column. The final product is a simulation tool with a focus on shallow water littoral acoustics, which can be used to define the sonar hardware and processing software necessary to meet various operational requirements.

An acoustic compliant coating is applied on a fluid-loaded structure to control the radiated pressure, by decoupling the fluid medium from the vibrating surface. In this thesis the problem of an infinite cylindrical shell immersed in a fluid and entirely covered with an acoustic compliant layer, excited either by a ring force or an incident acoustic plane wave is considered. To model this problem two different approaches are used. The first one, which is available in the literature, is based on multi-layer shell theory. In this approach the scalar and the vector potential formulation are used to solve for the response and the scattering from the cylinder. The second approach is based on modeling the compliant layer by a normally reacting impedance layer on the surface of the shell. The velocity response Green's function of the shell is found using the hybrid numerical/analytical method. Results for the radiated and scattered pressure from the shell are also presented. The advantage of this second approach is that it can be used to model complex coating geometries. The results obtained with both approaches are compared.

This thesis describes the design and implementation of an adaptive control system for active noise control. The main approaches available for implementing an active noise controller are presented and discussed. A Least Mean Squares (LMS) based algorithm, the Filtered-X LMS (FXLMS) algorithm, is selected for implementation. The significance of factors, such as delays, system output noise, system complexity, type and size of adaptive filter, frequency bandwidth, etc..., which can limit the performance of the adaptive control, is investigated in simulations. For hardware implementation, a floating-point DSP is selected to implement the adaptive controller. The control program and its implementation on the DSP are discussed. The program is first tested with a hardware-in-the-loop set-up and then implemented on a physical system. Active Noise Control in a duct is finally successfully demonstrated. The hardware and the results are discussed.

A numerical model that simulates the operation of a Forward Look Scan Sonar (FLSS) has been developed in this thesis. The model discretizes the sonar-projected signal by a set of rays using a geometrical approach. Bending of the rays due to varying acoustic wave speed is neglected. Simulated raw sonar data are generated, and used as input in the sonar processing algorithms to generate sonar images. Using the model, the influence of, the most critical characteristics of the sonar, including phase variations among the channels, non-homogeneous channel amplitude, and the number of bad channels, on the quality of the sonar image is determined. The results of the model are compared to real data from a low frequency FLS sonar (250 KHz) and a high frequency FLS sonar (600 KHz). There is good matching between the simulation and the operation of the two sonars and the performance was markedly enhanced by using the modeling results.

The vibrational and acoustic characteristics of fluid-loaded, cylindrical shells with single or multiple, aperiodically-spaced ring discontinuities are studied using an approach based on the mobility power flow (MPF) method and a hybrid numerical/analytical method for the evaluation of the velocity Green's function of the shell. The discontinuities are associated with internal structures coupled to the shell via ring junctions. The approach is a framework allowing alternative shell and/or internal structure models to be used. The solution consists of the net vibrational power flow between the shell and internal structure(s) at the junction(s), the shell's velocity Green's function, and the far-field acoustic pressure. Use of the MPF method is advantageous because the net power flow solution can be used as a diagnostic tool in ascertaining the proper coupling between the shell and internal structure(s) at the junction(s). Results are presented for two canonical problems: an infinite, thin cylindrical shell, externally fluid-loaded by a heavy fluid, coupled internally to: (1) a single damped circular plate bulkhead, and (2) a double bulkhead consisting of two identical damped circular plates spaced a shell diameter apart. Two excitation mechanisms are considered for each model: (1) insonification of the shell by an obliquely-incident, acoustic plane wave, and (2) a radial ring load applied to the shell away from the junction(s). The shell's radial velocity Green's function and far-field acoustic pressure results are presented and analyzed to study the behavior of each model. In addition, a comparison of these results accentuates the qualitative difference in the behavior between the single and multiple junction models. When multiple internal structures are present, the results are strongly influenced by inter-junction coupling communicated through the shell and the fluid. Results are presented for circumferential modes n = 0 & 2. The qualitative differences in the results for modes n = 0 and n = 2 (indicative of all modes n > 0ified in the far-field acoustic pressure and velocity Green's function response with the characteristics of the shell and internal plate bulkhead. The results for the single junction model demonstrate the significance of the shell's membrane waves on the reradiation of acoustic energy from the shell; however, when multiple junctions are present, inter-junction coupling results in a significant broad acoustic scattering pattern. Using the results and analysis presented here, a better understanding can be obtained of fluid-loaded shells, which can be used to reduce the strength of the acoustic pressure field produced by the shell.

The response of fluid-loaded plates has been extensively studied in the past.
However, most of the work deals with either infinite plates or finite plates with particular
boundary conditions and the results are generally presented only in the limit of small
wavelengths compared with the dimensions of the plates. Furthermore, the problem of
coupled finite plates where both the acoustic interaction and structural interaction are
included in the solution has not been considered. In this dissertation the response of two
coupled finite plates set in two alternative configurations is considered. The plates are
simply supported on two edges, with arbitrary boundary conditions on the remaining two
edges. The solutions obtained for the response of the plates include both the structural
interaction at the common junction and the acoustic interaction due to the scattered
pressure from each of the two plates. The results are presented in terms of the vibrational
power flow into and out of each plate component. The solution is based on a formulation developed in the wavenumber domain
combined with the Mobility Power Flow method. Using this approach, different
substructural elements coupled under different boundary conditions to form a complex
global structure can be considered. The detailed spatial and temporal scales of the structure response are not lost when using this method.
In obtaining the solution for the scattering from the fluid-loaded plates, a modal
decomposition in the direction normal to the simply supported edge is used. A spatial
Fourier-transform decomposition is used in the other direction. Due to the finiteness of
the plate, eight unknowns parameters are obtained in the transformed result. The solution
for these eight unknown parameters is obtained from the boundary conditions and the
condition that the response must remain finite. Two analytical approaches are used to
solve the final plate integral equation. The first approach consists of an approximation
method which obtains a solution based on the solution of the corresponding infinite plate
problem. The second approach is a more accurate solution based on the Projection
Method for the solution of integral equations.
Both of the approaches used in the solution provide accurate predictions at high
frequencies. At low frequencies especially for low structural damping or for heavy fluid
loading, only the Projection Method gives reliable results. This is attributed to the fact
that at low frequencies, the influence of the edges of the plates on the scattering is
significant.
The overall results obtained from this analysis indicate that the fluid loading and
the plate characteristics have a significant influence on the acoustic scattering properties,
especially in the case of heavy fluid loading.
The application of the method to coupled fluid-loaded plates indicates that the junction
enhances the scattering properties. The acoustical interaction between the coupled plates
increases the contribution to scattering from subsonic wavenumber components. In the
absence of the interaction, only supersonic wavenumbers contribute to the scattering.
Inclusion of acousticlal interaction requires both supersonic and subsonic components.
The significance of the contribution from the subsonic wavenumber components is
dependent on the type of the fluid loading.