Emotions are complex, universal experiences, which provide richness to life events influencing physiological and cognitive responses. For researchers exploring affect, there are a range of subjective reporting tools which allow an individual to evaluate their affective experience, subsequently allowing their physiological response to be tagged. However, the reliability and accuracy of these reports suffer from factors such as emotional intelligence, past emotional experiences and cultural background. To overcome these problems, this thesis presents the development of a novel tool, 'aPPRAISE', for obtaining implicit reports of affect. This tool allows an individual, whilst stimulated by music, to implicitly, continuously report their affective state upon a number of dimensions. In addition, aPPRAISE is statistically more reliable in tagging physiological data, than current popular tools. Furthermore, there are a number of dimensions available for researchers to assess induced affect e.g. valence, arousal, energy, tension and GEMS. It is generally accepted that two dimensions are not sufficient to describe the nuanced affective experiences elicited. However, obtaining reports on a large set of dimensions impacts experimental duration making the experiment laborious for participants. As a solution to these problems, the work presented in this thesis extracts the hidden relationship between dimensions of affect. Reliable models have been found which allow for a post-hoc estimation of unknown dimensions, such as energy, from a smaller set of reported dimensions, such as valence and arousal. To further emphasise the need of high-dimensional reports, the discrimination power of the GEMS space and alternate dimensional models are evaluated. The results highlight that a high-dimensional representation of affect significantly increases the ability to discriminate between affective experiences. Therefore for future research in this field, this thesis suggests the need to explore affect upon a large number of dimensions, using aPPRAISE and the models established for estimating un-reported dimensions of affect.

Advances in computer based optimization techniques can be used to enhance the efficiency of energy conversions processes, such as by reducing the aerodynamic loss in thermal power plant turbomachines. One viable approach for reducing this flow energy loss is by endwall contouring. This thesis implements a design optimization workflow for the casing geometry of a 1.5 stage axial flow turbine, towards mitigating secondary flow losses. In this thesis, a new non-axisymmetric endwall design method for the stator casing is implemented through a novel surface definition that draws from observations of the typical secondary flow pattern over the casing. The new casing design technique focuses on manipulating specific flow structures directly while also influencing the surrounding pressure field. This approach is tested on a three-dimensional axial turbine RANS model built in OpenFOAM Extend 3.2, with k-ω SST turbulence closure. Computer-based optimization of the surface topology is demonstrated towards automating the design process. This is implemented using Automated Process and Optimization Workbench (APOW) software. The designs are optimized using the total pressure loss across the full stage as the target function. The optimization and its sensitivity analysis give confidence that a good predictive ability is obtained by the Kriging surrogate model used in the prototype design process. The casing surface parametrization was shown to produce topologically smooth interfaces with the rest of the passage geometry. This was achieved by using the Beta distribution function to design a smooth casing groove path, which is a first application of the Beta distribution function to the contouring of a turbomachine casing. The flow analysis confirms the positive impact of the optimized casing groove design on the turbine isentropic efficiency compared to a reference diffusion based endwall design and compared to the benchmark axisymmetric design, at design and at off design conditions.

The aerodynamic noise radiating from an unsteady flow can be extracted by acoustic analogy from time-resolved Computational Fluid Dynamic (CFD) simulations. For this purpose, two Ffowcs Williams and Hawkings (FW-H) solvers are developed, based on an advanced time formulation (AFW-H) and on a convective formulation (CFW-H). The methods are coded in Python and embedded in Antares, a CFD post-processor of wide access and usability for the scientific community, developed by Cerfacs, France. The new FW-H solvers are tested on a hierarchy of noise sources of increasing complexity. The radiating field from elementary acoustic sources is considered first, progressing then to single-stream and dual-stream jets. The tests on monopoles, dipoles, and quadrupoles show good predictions of pressure fluctuation time-history and directivity against reference analytical results. CFD results obtained at Cerfacs by Large Eddy Simulation and at the University of Leicester by Detached Eddy Simulation provide the input to the acoustic analogy to estimate the noise radiation from jets. The jet noise predictions are compared against acoustic results obtained numerically by the elsA software (ONERA, France) and against sound measurements taken at the Von Karman Institute for Fluid Dynamics, Belgium. The tool is then used to assess dual-stream under-expanded jet noise in a configuration by Airbus SAS, at flow conditions that differ from the ones explored in previous aeroacoustic literature. Flight effects on jet noise are tested by applying the CFW-H tool to a single-stream under-expanded jet in-flight. The acoustic predictions for both static and in-flight jets are found in good agreement with reference predictions and with measurements, building confidence in using the new FW-H solvers to extract the aerodynamic noise generated by unsteady shock-containing jets.

As the cost of wind energy becomes of increasing importance to the global surge of clean and green energy sources, the reliability-critical power converter is a target for vast improvements in availability through dedicated research. To this end, this thesis concentrates on providing a new holistic approach to converter reliability research to facilitate reliability increasing, cost reducing innovations unique to the wind industry. This holistic approach combines both computational and physical experimentation to provide a test bench for detailed reliability analysis of the converter power modules under the unique operating conditions of the wind turbine. The computational models include a detailed permanent magnet synchronous generator wind turbine with a power loss and thermal model representing the machine side converter power module response to varying wind turbine conditions. The supporting experimental test rig consists of an inexpensive, precise and extremely fast temperature measurement approach using a PbSe photoconductive infra-red sensor unique in the wind turbine reliability literature. This is used to measure spot temperatures on a modified power module to determine the junction temperature swings experienced during current cycling. A number of key conclusions have been made from this holistic approach. -Physics-of-failure analysis (and indeed any wind turbine power converter based reliability analysis) requires realistic wind speed data as the temporal changes in wind speed have a significant impact on the thermal loading on the devices. -The use of drive train modelling showed that the current throughput of the power converter is decoupled from the incoming wind speed due to drive train dynamics and control. Therefore, the power converter loading cannot be directly derived from the wind speed input without this modelling. -The minimum wind speed data frequency required for sufficiently accurate temperature profiles was determined, and the use of SCADA data for physics-of failure reliability studies was subsequently shown to be entirely inadequate. -The experimental emulation of the power converter validated a number of the aspects of the simulation work including the increase in temperature with wind speed and the detectability of temperature variations due to the current's fundamental frequency. Most importantly, this holistic approach provides an ideal test bench for optimising power converter designs for wind turbine, or for other industries with stochastic loading, conditions whilst maintaining or exceeding present reliability levels to reduce wind turbine's cost of energy, and therefore, society.

This thesis investigates paper-based electronics in terms of various substrates, fabrication methods and example devices, including touch sensors and microwave resonators. The term ‘paper’ is very broad and covers a wide range of substrates. A decision matrix has been created to determine the optimum paper for an application, based on a range of properties. Thermal evaporation and screen printing are compared for their use as fabrication methods for paper-based electronics and a second decision matrix has been compiled. Based on these decision matrices, screen printing onto a thicker matt paper was determined to be optimal. The printing process was further optimised to achieve the best results from the in-house process. Using this well-developed screen-printing method, passive components (including inductors and interdigitated capacitive touch sensors) were fabricated and found to be comparable with state-of-the- art results reported in the literature. Measurements from the touch pads were compared to modelling, with little variation between the two, and were confirmed to work under a wide range of conditions, showing that they are compatible with any user. The microwave characteristics, up to 3GHz, of both the chosen substrate and silver-flake ink were investigated through production of screen-printed transmission lines. These characteristics were then used to create microwave resonators. The frequency range is important for applications as the industrial, scientific and medical radio band (ISM band) lies between 2.45 and 2.55 GHz which includes Wi-Fi and Bluetooth. Initially, stub resonators were considered to determine the cause of differences between theoretical and measured results. Then spiral defected ground structures were made, with multiple resonances, and sensitivity to touch and humidity demonstrated. As paper is hygroscopic, the effect of humidity on paper-based electronics is of key importance. This has been considered for all the devices fabricated in this work and it has been determined that the change in permittivity of the substrate, as a result of absorbed water within paper, is the most dominant factor.

Conventional element-based methods for crack modelling suffer from remeshing and mesh distortion, while the cracking particle method is meshless and requires only nodal data to discretise the problem domain so these issues are addressed. This method uses a set of crack segments to model crack paths, and crack discontinuities are obtained using the visibility criterion. It has simple implementation and is suitable for complex crack problems, but suffers from spurious cracking results and requires a large number of particles to maintain good accuracy. In this thesis, a modified cracking particle method has been developed for modelling fracture problems in 2D and 3D. To improve crack description quality, the orientations of crack segments are modified to record angular changes of crack paths, e.g. in 2D, bilinear segments replacing straight segments in the original method and in 3D, nonplanar triangular facets instead of planar circular segments, so continuous crack paths are obtained. An adaptivity approach is introduced to optimise the particle distribution, which is refined to capture high stress gradients around the crack tip and is coarsened when the crack propagates away to improve the efficiency. Based on the modified method, a multi-cracked particle method is proposed for problems with branched cracks or multiple cracks, where crack discontinuities at crack intersections are modelled by multi-split particles rather than complex enrichment functions. Different crack propagation criteria are discussed and a configurational-force-driven cracking particle method has been developed, where the crack propagating angle is directly given by the configuration force, and no decomposition of displacement and stress fields for mixed-mode fracture is required. The modified method has been applied to thermo-elastic crack problems, where the adaptivity approach is employed to capture the temperature gradients around the crack tip without using enrichment functions. Several numerical examples are used to validate the proposed methodology.

Current power system is in the transition from traditional power grid to Smart Grid. A key advantage of Smart Grid is its integration of advanced communication technologies, which can provide real-time system-wide two-way information links. Since the communication system and power system are deeply coupled within the Smart Grid system, it makes Quality of Service (QoS) performance analysis much more complex than that in either system alone. In order to address this challenge, the effective rate theory is studied and extended in this thesis, where a new H transform based framework is proposed. Various scenarios are investigated using the new proposed effective rate framework, including both independent and correlated fading channels. With the effective rate as a connection between the communication system and the power system, an analysis of the power grid observability under communication constraints is performed. Case studies show that the effective rate provides a cross layer analytical framework within the communication system, while its statistical characterisation of the communication delay has the potential to be applied as a general coupling point between the communication system and the power system, especially when real-time applications are considered. Besides the theoretical QoS performance analysis within Smart Grid, a new Software Defined Smart Grid testbed is proposed in this thesis. This testbed provides a versatile evaluation and development environment for Smart Grid QoS performance studies. It exploits the Real Time Digital Simulator (RTDS) to emulate different power grid configurations and the Software Defined Radio (SDR) environment to implement the communication system. A data acquisition and actuator module is developed, which provides an emulation of various Intelligent Electronic Devices (IEDs). The implemented prototype demonstrates that the proposed testbed has the potential to evaluate real time Smart Grid applications such as real time voltage stability control.

With increased energy demand and decreased fossil fuels usages, the penetration of distributed generators (DGs) attracts more and more attention. Currently centralized control approaches can no longer meet real-time requirements for future power system. A proper decentralized control strategy needs to be proposed in order to enhance system voltage stability, reduce system power loss and increase operational security. This thesis has three key contributions: Firstly, a decentralized coordinated reactive power control strategy is proposed to tackle voltage fluctuation issues due to the uncertainty of output of DG. Case study shows results of coordinated control methods which can regulate the voltage level effectively whilst also enlarging the total reactive power capability to reduce the possibility of active power curtailment. Subsequently, the communication system time-delay is considered when analyzing the impact of voltage regulation. Secondly, a consensus distributed alternating direction multiplier method (ADMM) algorithm is improved to solve the optimal power ow (OPF) problem. Both synchronous and asynchronous algorithms are proposed to study the performance of convergence rate. Four different strategies are proposed to mitigate the impact of time-delay. Simulation results show that the optimization of reactive power allocation can minimize system power loss effectively and the proposed weighted autoregressive (AR) strategies can achieve an effective convergence result. Thirdly, a neighboring monitoring scheme based on the reputation rating is proposed to detect and mitigate the potential false data injection attack. The simulation results show that the predictive value can effectively replace the manipulated data. The convergence results based on the predictive value can be very close to the results of normal case without cyber attack.

This thesis describes the development of a new seal concept for steam turbines called the Aerostatic Seal. The Aerostatic Seal is a dynamic seal, and so can respond to rotor radial movement to maintain a low clearance between the seal and the rotor. As the seal is dynamic, smaller clearances can be achieved without rotor contact compared to conventional static seals such as the labyrinth seal, hence increasing the efficiency of the turbine through reduced leakage. Furthermore, as the seal is dynamic it can tolerate larger radial transients typically found during start up and shut down of the steam turbine, and so also contributes to increasing the flexibility of the turbine plant. In this thesis an analytical design and analysis methodology was developed for the Aerostatic Seal. The methodology was used to generate a number of seal designs which were experimentally tested in a non-rotating test facility using room temperature air. The results confirmed that the seal would operate dynamically, and the experimental campaign provided valuable data on the operation of the seal. The non-rotating rig was also used to test a second generation seal design. The seal was then tested in a rotating test facility, which modelled high speed rotor radial transients with an adjustable eccentric rotor. The Aerostatic Seal demonstrated the ability to respond to high speed transients. A final test campaign was conducted in the high temperature steam rig at TU Braunschweig, Germany, enabling experimental demonstration of the Aerostatic Seal using realistic materials and represented realistic steam turbine conditions. Finally, based on the experimental and analytical work carried out within this thesis, a proposed Aerostatic Seal design for steam turbine implementation is presented.

The ability to correctly capture large deformation behaviour in solids is important in many problems in geotechnical engineering such as slope failure or installation of foundations. The Material Point Method (MPM) is a computational method with particular suitability for modelling problems involving large deformations. In the MPM, a domain is modelled using a set of material points at which state variables are stored and tracked. These material points move through a fixed background grid upon which calculations take place with variables being mapped between the material points and the grid. This thesis sets out to develop the MPM as a method with potential for use in geotechnical problems. Problems are encountered with the original MPM when material points cross between grid cells, and one solution to this is the Generalised Interpolation Material Point (GIMP) method, where material points are able to influence nodes beyond the currently occupied grid cell. Most development of the GIMP method has used an explicit approach, however there are a number of advantages of an implicit approach including larger load steps and improved error control. This thesis focuses on the development of a large deformation elasto-plastic implicit GIMP method. A way of calculating the deformation gradient consistent with the MPM is introduced and convergence is demonstrated using this method which has previously been frequently omitted from MPM research. An alternative way of updating material point domains using the stretch tensor is also proposed. The MPM has a number of similarities to the FEM, and it is often suggested that FEM technologies are trivial to use with the MPM. The MPM can encounter localisations caused by shear banding and, to overcome this, a gradient plasticity approach previously implemented for the FEM is investigated with the GIMP method for the first time. The addition of gradient plasticity to the GIMP method introduces a length scale parameter which governs the width of these shear bands and removes the mesh dependency which is encountered with conventional approaches. It is shown that implementation is possible however, there are a number of problems that are present in the combination of the two methods which should not be overlooked in the future.