Structures made of plate-like components are common in a variety of industries where the impacts of structural failures are severe. In many cases these structures are surrounded and only partially accessible, such as storage tanks and bridges, making them difficult to inspect frequently. The application of ultrasonic Guided Waves (GWs) in the evaluation and monitoring of relatively large plate-like structures is evermore a feasible option with the continuous development of transducer arrays. The use of transducer arrays is however complex due to directional control and the existence of many GW modes. Aimed at the evaluation of plate-like structures, in this research two piezoelectric transducer arrays respectively capable of omnidirectional and unidirectional control of the fundamental GW shear mode in plates (SH0) with above 20 dB mode purity are successfully designed, produced and validated. Omnidirectionality facilitates full structural evaluation coverage and can lead to defect mapping of large volumes with relatively few transducers. A unidirectional beam with relatively high mode purity facilitates evaluation of specific structural locations. Preference to the SH0 mode was given because of its non-dispersive and in-plane propagation properties making it more suitable than other GW modes to propagation in structures surrounded by fluid material. To enable the array development, a number of monolithic piezoelectric thickness-shear transducers of varied area were characterised with respect to GW mode directionality, amplitude and SH0 mode purity. The characterisation of each thickness-shear transducer allows for optimised superposition manipulation for specific applications. A single characterised shear transducer was selected for use in the development of omnidirectional and unidirectional SH0 mode transducer arrays. To aid development a linear superposition analysis model was produced and used to predict for a circular array design the optimum parameters for omnidirectional SH0 mode transmission with significant mode purity. A range of parameter combinations were evaluated and their predicted influence on array performance was characterised. The same method was employed to optimise a dual row linear array design for the unidirectional transducer array. All results were validated by FE models and later with empirical data. Both developed transducer arrays were characterised with respect to GW mode directionality, magnitude and SH0 mode purity. Both their detection sensitivity to pertinent defects and structures was validated, demonstrating relevance to Non Destructive Evaluation (NDE) and Structural Health Monitoring (SHM) applications.

This thesis investigates the broad subject of thermal management problems currently encountered in Formula One race car design. A computationally economical tool, based on linear superposition, for predicting the temperature field arising from a set of thermal and inlet velocity boundary conditions was developed. Using a set of base analyses, the research showed that it is possible to superpose and scale these results in order to predict the temperature field for differing sets of boundary conditions. This method was shown to have a significant speed advantage over typical computational simulations. An experimental facility was designed and built to provide validation for aspects of the linear superposition approach. A method of measuring the cylinder wall heat flux has been developed using thin film gauge technology. The resulting sensor was designed to fit the mounting of existing instrumentation in order to avoid requiring large scale modifications to existing test facilities. The design makes use of modern rapid prototyping techniques in order to meet this mounting requirement and to provide a novel solution to routing the signal from the thin film gauge. In addition, the research investigated a method for predicting the cylinder wall temperature in real-time. The cylinder wall is subject to heat fluxes from in-cylinder gases during the engine cycle on the inner face and the effect of the coolant jacket on the outer face. Two separate methods were used to process these thermal boundary conditions respectively, before being superposed in order to form the whole solution. The computation time of the method is characterised in order to demonstrate its feasibility for real-time operation.

In this thesis, we attempt to obtain a class of generalized bilinear differential equations in (3+1)-dimensions by Dp-operators with p = 5, which have resonant solutions. We construct resonant solutions by using the linear superposition principle and parameterizations of wave numbers and frequencies. We test different values of p in Maple computations, and generate three classes of generalized bilinear differential equations and their resonant solutions when p = 5.

No description available.

No description available.