The surface acoustic wave properties of Berlinite (a-AlPO4) have been investigated theoretically and experimentally, for a variety of crystallographic orientations, to evaluate its possible use as a substrate material for temperature compensated surface acoustic wave devices. A computer program has been developed to calculate the surface wave properties of a material from its elastic, piezoelectric, dielectric and lattice constants and their temperature derivatives. The program calculates the temperature coefficient of delay, the velocity of the surface wave, the direction of power flow and a measure of the electro-mechanical coupling. These calculations have been performed for a large number of orientations using a modified form of the data given by Chang and Barsch for Berlinite and predict several new temperature compensated directions. Experimental measurements have been made of the frequency-temperature response of a surface acoustic wave oscillator on an 80° X axis boule cut which show it to be temperature compensated in qualitative agreement with the theoretical predictions. This orientation shows a cubic frequency-temperature dependence instead of the expected parabolic response. Measurements of the electro-mechanical coupling coefficient k gave a value lower than predicted. Similar measurements on a Y cut plate gave a value which is approximately twice that of ST cut quartz, but again lower than predicted. The surface wave velocity on both these cuts was measured to be slightly higher than predicted by the computer program. Experimental measurements of the lattice parameters a and c are also presented for a range of temperatures from 25°C to just above the alpha-beta transition at 584°C. These results are compared with the values obtained by Chang and Barsch. The results of this work indicate that Berlinite should become a useful substrate material for the construction of temperature compensated surface acoustic wave devices.

The sound field in rooms of small dimensions used for music reproduction is characteristically different from that found in larger rooms for music performance such as auditoria. Key differences between small critical listening spaces and large auditoria are the vastly different ranges of energy decay, 100 ms for the former and up to 8 s for the latter, and its directional behaviour, typically non-diffuse for the former and approximating a diffuse field for the latter. Despite these substantial differences, most of the metrics developed to describe the sound field in large spaces are evoked to quantify the performance of small rooms. This project focuses on developing measurement methods to characterise temporal and spatial qualities of sound in small rooms. A number of methods based on currently available acoustic probes have been developed. The implementation requisites and accuracy for each method has been quantified. Factors such as direction, time of arrival and strength of reflections have been extracted using signal analysis techniques based on the active instantaneous intensity and short-time Fourier transform. These factors are subsequently mapped to allow a description of their evolution through the energy decay in the room for a given measurement location. The best performing system, based on the use of one-dimensional p-p intensity probe mounted in a custom cradle, achieves a minimum overall mean error of 0.226 degrees and 2.971 degrees for the direct sound and first reflection respectively, which is near or below the measured human minimum audible angle (MAA). The method developed has direct applications in the quantification of small room acoustic sound fields for critical listening purposes.

Soundfield reproduction has a vast array of applications, yet the ecological validity – and external validity – has been little researched beyond the narrow limits of the physical approach. This mixed model research explored the differing perception of reproduced urban soundfields and urban soundscapes. The methodology analysed the influences of: auralisation technique; soundscape; place affordances; time; space; and memory – subject agency and expertise. Three methods of data collection and analysis were employed: a semantic differential analysis and two psycholinguistic methods – one existing and one novel. The semantic differentials’ soundscape dimensions of the public’s response from in situ listening were replicated by participants’ responses from ambisonic reproduction but not binaural. A generalised soundscape dimension model was proposed that integrates the structural model of appraisal theory with the interpretation of motivation-affordance fit and mediation dimensions. Different soundscapes were evaluated differently and stimulated differing processes of perception, which in turn effected reproducibility. A focus group was used as well as members of the public and laboratory participants. It was found that experts responded more in terms of source identification. Existing methods found no significant difference between in situ and ambisonic listening. The novel method found the ambisonic soundfield was described in terms of objects in motion or sound objects whilst the binaural soundfield and in situ soundscape were described in terms of sources or activity. As an assessment of external validity, the novel psycholinguistic method found that binaural reproduction held validity over ambisonic reproduction. An in situ real-time binaural reproduction test sought to isolate the ‘electroacoustic ear’ – the findings were consolidated and discussed in terms of embodiment and ventriloquism. The novel psycholinguistic analysis provided a more accurate representation of the cognitive process of soundscape perception and is offered as a tool for the external validity assessment of urban soundfield reproduction.

The research topic of this thesis is the use of inverse filtering for the design and implementation of two-channel virtual acoustic imaging systems that utilise loudspeakers. The basic objective of such systems is to invert the electroacoustic plant between the input to the loudspeakers and the output at the listener’s ears and hence make it possible for a pair of binaural signals to be locally reproduced at the position of the listener’s ears. As a starting point for the research presented, a previously introduced type of inverse filtering design is considered in which the inverse is implemented with FIR filters. The basic formulation of this design is described and a number of innovative points regarding its implementation are made. An experimental procedure is then formulated for the evaluation of the effectiveness of this inverse filtering design that is based on objective measurements of the inversion process. Unlike previously employed methods that are based on computer simulations or subjective experiments, the introduced experimental procedure is shown to be very efficient in isolating and exactly quantifying the effect on the accuracy of the inversion of a number of errors and approximations typically present in the implementation. A detailed evaluation is thus presented of the inverse filtering design at hand in realistic conditions of implementation. Subsequently, a novel method for the off-line implementation of the inverse filtering is presented that utilises recursive filters of lower order. In this method, the responses of the inverse filters are decomposed into two parts, one realisable in forward time and one in backward time. The effectiveness of this new method for the implementation of the inverse is tested and compared with a small selection of the objective evaluation results described above. Finally, an algorithm for the on-line implementation of the forward-backward inverse filtering is proposed and its computational cost is compared with the currently available frequency-domain block-processing filtering algorithms.

Sound visualisation techniques have played a key role in the development of acoustics throughout history. Progress in measurement apparatus and the techniques used to display sound and vibration phenomena has provided excellent tools for understanding specific acoustic problems. Traditional methods, however, such as step-by-step measurements or simultaneous multichannel systems, require a significant trade-off between time requirements, flexibility, and cost. This thesis explores the foundations of a novel sound field mapping procedure. The proposed technique, Scan and Paint, is based on the acquisition of sound pressure and particle velocity by manually moving a p-u probe (pressure-particle velocity sensor) across a sound field, whilst filming the event with a camera. The sensor position is extracted by applying automatic colour tracking to each frame of the recorded video. It is then possible to directly visualise sound variations across the space in terms of sound pressure, particle velocity or acoustic intensity. The high flexibility, high resolution, and low cost characteristics of the proposed measurement methodology, along with its short time requirements, define Scan and Paint as an efficient sound visualisation technique for stationary sound fields. A wide range of specialised applications have been studied, proving that the measurement technique is not only suitable for near-field source localisation purposes but also for vibro-acoustic problems, panel noise contribution analysis, source radiation assessment, intensity vector field mapping and far field localisation.

This thesis presents a study of acoustic cavitation generated in an ultrasonic reactor, with the particular aim of enhancing hydrogen gas production and release. The stabilisation of cavitation clusters formed in a set of ultrasonic reactors is demonstrated. The highly stable cluster is induced by the positioning of a rod at the antinode of the sound field employed. These sound fields were characterised with a new technique based on particle imaging. Here adding rheoscopic particles within such reactors revealed a novel and useful method for the characterisation of standing wave fields. This observation was supported by pressure measurements using a hydrophone. In addition the stabilised cluster was investigated using an electrochemical method to monitor the erosion of the surface directly above the cluster, at short (10’s of μs) and long (100’s of s) timescales. Both timescales indicate changes in the stability and nature of the cluster, which in turn is dependent on the local surface conditions (roughness) of the rod/electrode assembly. Low light level imaging of the stabilised cavitation cluster demonstrates the occurrence of sonochemiluminescence (SCL). It is shown that the spatial extent of light emitted via SCL is correlated with the pressure amplitude of the sound field. A visual ‘shimmer’ effect is also shown to be emanating from stabilised cavitation clusters. This is attributed to local heating which in turn induces refractive index changes, which are enhanced through the use of Schlieren imaging. This local cluster induced-heating of the liquid is quantified using a variety of physical measurements. Investigation into the ultrasonic enhancement of the production of molecular hydrogen from aluminium corrosion is made. This study showed that the sonochemical enhancement was insignificant compared to local heating effects associated with the sound field. Analysis of the performance of an electrolysis system, designed and manufactured by the project sponsors (HTOGO Ltd.), is reported. Measurement of the hydrogen gas produced by the system highlights a low Faradaic efficiency and long response time for gas release. An innovative method for the rapid release of gas via ultrasonic outgassing of a liquid reservoir, containing hydrogen and oxygen gas bubbles, is demonstrated. A novel optically isolated Coulter counter system for the in-situ determination of the size distribution of bubbles in a bubbly liquid reservoir is reported. This thesis illustrates the underpinning principles of this technique and the determination of the best calculation method for successful calibration and accurate measurement of the bubble size distributions generated in an electrochemical reactor.

The efficiency of travelling wave thermoacoustic system, for a given operating temperature difference, is determined by the acoustic wave transmission through its feedback resonator loop system. Curved duct bends are one of the most repeated components used in the construct of these feedback resonator systems and thus require acoustic transmission optimisation. This research investigates the transmission of low frequency, high amplitude acoustic waves propagating through duct bends with different radius of curvatures using Particle Image Velocimetry (PIV). The experimental PIV investigation was conducted on the axial plane of the bend. The velocity vector maps obtained from each run was analysed using both the newly developed Velocity based Wave Decomposition (VWD) technique as well as the Proper Orthogonal Decomposition (POD) technique. The POD technique was shown to successfully separate the different flow component of the acoustic wave in the respective Proper Orthogonal Modes (POMs). The acoustic transmission was thus computed based on the strength of these POMs. The POMs also allowed for the flow visualisation of the different loss mechanism that exists within the wave propagating through the bend (most notably, the energy cascade loss mechanism). Based on the quantitative measurement of the acoustic transmission as well as the qualitative flow observation of the different loss mechanism, a non-dimensional parameter was developed in order to characterise the acoustic transmission through curved duct bend systems. This parameter is known as the Strouhal-Dean number. Based on this parameter, the acoustic transmission can be characterised into 3 acoustic flow regimes: Viscous dominated oscillation, Inertia dominated oscillation and the Transition regime between the first two oscillation regimes. The optimum acoustic transmission range corresponded to the transition regime where the inertia generated secondary circulation was balanced by its viscous loss suppression. The optimal Strouhal Dean number for acoustic transmission was found to be approximately 10.8.

Combining modes of vibration, such as longitudinal and torsional vibration, is advantageous in many ultrasonic applications such as ultrasonic drilling, welding, and motors. In this work we present a novel approach to the design a longitudinal-torsional (LT) ultrasonic horn which adapts the front mass in a traditional Langevin transducer. Different approaches, such as degeneration of longitudinal vibration and coupling between longitudinal (L) and torsional (T) modes, have been used to generate the LT mode of vibration. The degeneration approach creates a non-uniform section, by cutting and twisting a number of slots along the path of the L wave such that part of the wave converts into T wave whilst the remaining part propagates unchanged through the section; these two parts are recombined near the output surface to form LT vibration. The mode coupling approach uses two set of vibration generators, usually piezoelectric elements, where one set generates L vibration whilst the second set generates T vibration. An exponential cross-sectional horn uses to combine the two modes where the area reduction factor is selected such that these modes resonate at the same frequency. However, many limitations prevent the wide usage of these methods in ultrasonic applications. These limitations are the complex design and excitation, possible coupling with surrounding modes, instability in operating at different boundaries, difficulty in securing the structure without influencing the vibrational response and the low produced torsionality, which is the ratio of torsional to longitudinal response at the output face. The new approach is based on combining the principles of these methods to overcome the previously stated limitations, the slotting technique is incorporated into the exponential cross-sectional path and the horn produced is utilised as the front mass of a Langevin transducer. A set of design and performance criteria are used to optimise the transducer and includes applicable design; methods of securing the transducer; and the excitation features of LT transducer such that it can operate without the effects of surrounding modes of vibration and can produce high response and torsionality at the output surface. A methodology which combines mathematical and experimental modelling is used to optimise LT transducer design. The mathematical modelling, which includes finite element (FE) and analytical methods, is performed to optimise the geometry and to predict electromechanical parameters, modal parameters and the dynamic behaviour of LT transducer. The experimental modelling is used to validate the mathematical results and to characterise the fabricated prototypes under different operating conditions. The dimensions of the initial design of the L mode Langevin transducer are derived from the principles of the wave equation. This transducer has a set of piezoceramic components sandwich between a cylindrical back mass and an exponential front mass connected by a pre-stressed bolt. The dimensions are used to create the FE model, using the FE software package ABAQUS, where different shapes of cut at different dimensions and at various angle of twist along the front mass are introduced and examined by a modal analysis procedure to the front mass. An optimised model is then utilised in a size scaling study to confirm the suitability of using this approach for different ultrasonic applications. The dimensions of the optimised design are also used in the analytical study, based on Mason’s electric equivalent circuit approach, to predict the electromechanical parameters where a one-dimensional equivalent circuit approach is created separately for each part whilst the combination vibrational motion in the front mass is represented by two, longitudinal and torsional, equivalent circuits. The complete equivalent network of the LT transducer is then solved using the mathematical software package MATHEMATICA. The analytical model is also extended to validate some of particular FE findings such as the distribution of the response amplitude and the location of the longitudinal nodal plane along the transducer’s structure. Two optimised models of different sizes are fabricated and characterised through different testing techniques including electrical impedance analysis, experimental modal analysis (EMA) and experimental harmonic analysis. Optimisation of the pre-stressing of the transducer is performed by applying different torques to the pre-stressed bolt and measuring the electrical impedance spectra where the results are compared to analytical findings. EMA is then used to describe the natural characteristics of the structures where the results are used to accurately extract the modal parameters and to validate the predictions of the FE and analytical model. Different levels of harmonic excitation are used to characterise the fabricated prototypes where the results are compared to the findings of the mathematical modelling. A case study of the design of the LT drill is presented to validate the design approach for real ultrasonic applications. A similar methodology is applied and the resulting LT drill is tested for both unloaded and loaded operating conditions. The results obtained show that this new approach can be easily and successfully applied to ultrasonic applications to produce a torsional to longitudinal amplitude response of 0.8 which is measured on a fabricated prototype.

Humpback whales songs have been widely investigated in the past few decades. This study proposes a new approach for the classification of the calls detected in the songs with the use of Hidden Markov Models (HMMs). HMMs have been used once before for such task but in an unsupervised algorithm with promising results. Here HMMs were trained and two models were employed to classify the calls into their component units and subunits. The results show that classification of humpback whale songs from one year to another is possible even with limited training. The classification is fully automated apart from the labelling of the training set and the input of the initial HMM prototype models. Two different models for the song structure are considered: one based on song units and one based on subunits. The latter model is shown to achieve better recognition results with a reduced need for updating when applied to a variety of recordings from different years and different geographic locations.

Bowed instruments are among the most exciting sound sources in the musical world, mostly because of the expressivity they allow to a musician or the variety of sounds they can generate. From the physical point of view, the complex nature of the nonlinear sound generating mechanism – the friction between two surfaces – is no less stimulating. In this thesis, a physical modelling computational method based on a modal approach is developed to perform simulations of nonlinear dynamical systems with particular application to friction-excited musical instruments. This computational method is applied here to three types of systems: bowed strings as the violin or cello, bowed bars, such as the vibraphone or marimba, and bowed shells as the Tibetan bowl or the glass harmonica. The successful implementation of the method in these instruments is shown by comparison with measured results and with other simulation methods. This approach is extended from systems with simple modal basis to more complex structures consisting of different sub-structures, which can also be described by their own modal set. The extensive nonlinear numerical simulations described in this thesis, enabled some important contributions concerning the dynamics of these instruments: for the bowed string an effective simulation of a realistic wolf-note on a cello was obtained, using complex identified body modal data, showing the beating dependence of the wolfnote with bowing velocity and applied bow force, with good qualitative agreement with experimental results; for bowed bars the simulated vibratory regimes emerging from different playing conditions is mapped; for bowed Tibetan bowls, the essential introduction of orthogonal mode pairs of the same family with radial and tangential components characteristic of axi-symmetrical structures is performed, enabling an important clarification on the beating phenomena arising from the rotating behaviour of oscillating modes. Furthermore, a linearized approach to the nonlinear problem is implemented and the results compared with the nonlinear numerical simulations. Animations and sounds have been produced which enable a good interpretation of the results obtained and understanding of the physical phenomena occurring in these system.