In 2004 the Indian Ocean tsunami once again showed the world the destructive capabilities of these natural disasters. In this work this practical problem was abstracted to a simplified problem of waves propagating over complex geometry, specifically the processes involved in a solitary and leading depression wave interacting with a bottom seated semi-cylinder over uniform depth and shoaling topography. A purpose written Arbitrary-Lagrangian-Eulerian finite element code to solve the two-dimensional Navier-Stokes equation was developed. It utilised high performance libraries PETSc, Hypre and Triangle to perform simulations with approximately 8 million degrees of freedom across 256 cores on UCL’s supercomputer, Legion. The code was validated against five benchmark test cases and mesh convergence was shown. The code exhibited linear weak scaling and superlinear strong scaling and with this capability, these problems could be analysed using two tools; firstly, global conservation measures such momentum and energy. The rate of change of these integral measures were related to forces on obstacles and boundaries and agreement with model estimates was found even when the wave/boundary interactions were complex. Secondly, flow diagnostics such as the rate of strain and vorticity were used to characterise the velocity field. The large scale capability also allowed the free surface boundary layer to be captured which revealed an (weaker) oscillatory nature of that found at a rigid boundary. The code could not handle breaking and inundation, therefore experiments were carried out to study a depression wave as it interacts and breaks on a beach. Using a novel mechanism for generating isolated depression waves, general features such as extent and speed of recession and surge where characterised and it is noteworthy that the shoreline recession, a defining feature of shoaling depression waves, which was seen during the 2004 tsunami, was captured.

Microbubbles coated with polymers or surfactants have been used in medical imaging for several years as ultrasound contrast agent particles and are now being investigated by researchers as drug and gene delivery vehicles and blood substitutes. Current methods available for the preparation of microbubbles are insufficient as they result in microbubbles with a wide size distribution and as such filtration is necessary before their use. With a view to fill the above demand, a detailed investigation has been carried out in this research to learn the viability of co-axial electrohydrodynamic atomization (CEHDA) technique to prepare microbubbles. The research also focuses on the effects of the process parameters such as flow rates, applied voltage and material parameters such as electrical conductivity, surface tension and viscosity with the objective of preparing polymer or surfactant coated stabilized microbubbles with diameters < 8 μm and with a narrow size distribution. A model glycerol-air system was used so that the CEHDA technique was modified to generate suspensions of microbubbles to a diameter < 8 μm with a narrow size distribution and then to characterise the CEHDA microbubbling process in terms of size and stability with varying process parameters and material parameters. Construction of a parametric plot between the air flow rate and the liquid flow rate was extremely useful in identifying the flow rate regime of air and liquid or suspension or solution for the continuous microbubbling of the system used. With further investigations into the CEHDA microbubbling technique, it was possible to develop strategies, first, to prepare suspensions of stabilized phospholipids-coated microbubbles with a mean diameter of ~ 5 μm and a polydispersivity index of 9%, and second, polymeric microspheres with a mean diameter of 400 nm and a polydispersivity index of 8% using a biocompatible polymer.

The current study presents the modelling of turbulent non-premixed or premixed flames over a range of combustion regimes using a variety of fuel mixtures. A parabolic Finite Volume method is used for the flow solution coupled to a joint-scalar transported Probability Density Function (PDF) approach for the inclusion of the thermochemistry without approximation. Finite chemistry effects were studied for two different cases. Moderate or Intense Low Oxygen Dilution (MILD) combustion is quantified with excellent pollutant formation, showing a reduction in temperature gradients and an increasing distributed reaction zone with dilution. High shear flows with low Damköhler numbers are also investigated with the appearance of a neck zone and distributed reaction at higher jet velocities. Transported PDF methods produce good agreement with experimental results, where discrepancies in the mixing model and flow field characterisation are apparent. Molecular mixing is closed using the modified Curl's model which provides reasonable mixing behaviour. Further studies into improvements upon the micro-mixing model are encouraged. Additionally, sensitivity to boundary conditions is demonstrated. The cases studied contribute to the understanding of emerging trends in practical combustion devices and portray finite-chemistry effects such as extinction and re-ignition.

Network Rail records indicate that approximately 24% of the total track maintenance and renewal budgets are spent on railway switches and crossings (S&C), which account for only 5% of the total main line track mileage. S&C complexities also introduce a degree of risk, which must be adequately managed to ensure a safe and reliable network. In recent years, risk mitigation fell short, resulting in some high profile incidents at S&C. A recent derailment investigation uncovered knowledge gaps within the UK rail industry, including the understanding of S&C degradation. This PhD research project was therefore initiated to investigating modelling tools for S&C wheel-rail interaction and degradation. A new wheel-rail contact detection routine has been developed and validated using existing software and a novel experimental technique using thermal imagery. Existing techniques were then integrated to enable the prediction of normal and tangential contact stresses whilst also simulating wear accumulation. To improve accuracy for long-term S&C damage, a combined tool for assessing non-Hertzian normal contact stresses and multiple modes of S&C degradation was sought. A novel 2.5D boundary element model capable of simulating wheel-rail contact detection, surface and sub-surface elastic and elastic-plastic stress analysis and dynamic material response is presented. Superior computational effort is also achieved, illustrating further the feasibility of such an approach. To conclude, a three-dimensional dynamic finite element model of a railway wheel passing through a cast manganese crossing has also been developed. For the first time, a tool capable of simulating both dynamic contact forces and corresponding plastic material response has been used to discover flaws within existing designs of UK cast manganese crossings. This approach has enabled immediate recommendations for asset improvement to be provided to Network Rail and gives the UK rail industry more scientific insight into the optimal design of railway crossings.

Soot particles formed and emitted from (e.g.) direct-injected diesel engines are dangerous to human health and legislative measures used to reduce emis- sions pose a technical challenge for manufacturers. Models of soot formation and oxidation may therefore be useful tools for developing engines and con- trol strategies. In the present work, a sectional soot model able to reproduce the soot particle size distribution (PSD) is applied to laminar premixed and diffusion flames as well as a reactor system. The soot PSDs in laminar premixed stagnation flow flames were found to be sensitive to the coagula- tion collision efficiency and a novel model was developed. The soot model under-predicted soot volume fraction levels when applied to a set of laminar ethylene and propane counter-flow diffusion flames and a sensitivity anal- ysis suggested further assessment of the formation of poly-cyclic aromatic hydrocarbons (PAH) was needed. The chemical reaction mechanism was subsequently assessed using species measurements from a laminar premixed benzene flame and selected parts of the reaction mechanism reviewed. Rea- sonable agreement was obtained, including for formation of PAHs. However, non-existing or insufficient oxidation paths of some PAH species, including pyrene, may contribute to over-predictions by the soot model during non- sooting conditions. Formation of PAHs in a laminar ethylene counter-flow diffusion flame was investigated next. The agreement between calculations and measurements was found to be reasonable for major, minor and single ring aromatic species. However, the calculated concentrations of all PAH species are under-predicted. The under-prediction of pyrene is comparable to the under-prediction of the soot volume fraction in some of the diffusion flames previously investigated, making the uncertainty of the PAH chem- istry a possible explanation. Future soot modelling research should therefore focus on investigating the PAH chemistry for different types of flames and fuels.

It is commonly known that the turbocharger turbine is still designed using the quasi-steady assumption despite its highly pulsating unsteady working conditions. The positioning of a turbocharger in close proximity to the exhaust valve in order to extract substantial energy ultimately necessitates a thorough investigation regarding its performance under pulsating flow conditions. This thesis presents experimental and numerical work, as well as the design of new advanced stator concept to improve turbine performance under pulsating flow conditions. A cold flow test facility is setup mainly to isolate the effect of pulsating flow conditions and therefore allowing the performance deviation from the quasi-steady approach to be properly recorded and documented. Since experimental data alone is not sufficient for understanding the detailed flow field within the turbocharger turbine stage, a complete 3-D Computational Fluid Dynamics model is developed using commercial software Ansys CFX. The model is validated against experimental data for all steady and pulsating conditions. During pulsating conditions, the incidence angle close to the rotor inlet changed significantly which directly affected the turbine performance. A study on the turbine performance improvement by aggressive reduction of nozzle vanes are conducted and experimentally tested. Results of steady and pulsating conditions suggested that the new vanes arrangement delivered significantly improved performance under both operating conditions especially at 50% speed (equivalent to 30000 rpm). At 80% speed (48000 rpm), the turbine efficiency is either similar or better (up to 8 efficiency point improvement) than the baseline arrangements.

Surgical robots have proliferated in recent years, with well-established benefits including: reduced patient trauma, shortened hospitalisation, and improved diagnostic accuracy and therapeutic outcome. Despite these benefits, many challenges in their development remain, including improved instrument control and ergonomics caused by rigid instrumentation and its associated fulcrum effect. Consequently, it is still extremely challenging to utilise such devices in cases that involve complex anatomical pathways such as the spinal column. The focus of this thesis is the development of a flexible robotic surgical cutting device capable of manoeuvring around the spinal column. The target application of the flexible surgical tool is the removal of cancerous tumours surrounding the spinal column, which cannot be excised completely using the straight surgical tools in use today; anterior and posterior sections of the spine must be accessible for complete tissue removal. A parallel robot platform with six degrees of freedom (6 DoFs) has been designed and fabricated to direct a flexible cutting tool to produce the necessary range of movements to reach anterior and posterior sections of the spinal column. A flexible water jet cutting system and a flexible mechanical drill, which may be assembled interchangeably with the flexible probe, have been developed and successfully tested experimentally. A model predicting the depth of cut by the water jet was developed and experimentally validated. A flexion probe that is able to guide the surgical cutting device around the spinal column has been fabricated and tested with human lumber model. Modelling and simulations show the capacity for the flexible surgical system to enable entering the posterior side of the human lumber model and bend around the vertebral body to reach the anterior side of the spinal column. A computer simulation with a full Graphical User Interface (GUI) was created and used to validate the system of inverse kinematic equations for the robot platform. The constraint controller and the inverse kinematics relations are both incorporated into the overall positional control structure of the robot, and have successfully established a haptic feedback controller for the 6 DoFs surgical probe, and effectively tested in vitro on spinal mock surgery. The flexible surgical system approached the surgery from the posterior side of the human lumber model and bend around the vertebral body to reach the anterior side of the spinal column. The flexible surgical robot removed 82% of mock cancerous tissue compared to 16% of tissue removed by the rigid tool.

Artificial joints restore function which has been impaired due to disease, trauma or genetic conditions. Although many prostheses function satisfactorily for the lifetime of the patient, in some cases failure occurs due to wear of the bearing surfaces. To assess the wear life of artificial joint systems extensive testing in simulators is performed, yet there are discrepancies between the predicted performance and wear experienced in vivo. The lubricant used in simulators has a different chemical composition to synovial fluid. As the composition affects the performance of a lubricant, specifying a model fluid closer to synovial fluid will improve simulator testing, and consequently joint design. The difficulty in specifying a model fluid is that it is not fully understood which elements of synovial fluid determine the lubrication performance. This thesis aims to develop an understanding of how model synovial fluids can provide boundary lubrication in artificial joints. Static and dynamic measurements are used to reconcile the adsorption and lubrication behaviour of synovial fluid proteins and lipids. The boundary lubrication of both current prosthesis materials and cutting-edge artificial cartilage materials is considered. Static tests show changes to adsorbed films dependent on surface interactions, pH and interactions between proteins and lipids. This affects the hydration and macromolecular content of adsorbed layers, which impacts on the layer properties. Under loading and shear, pH and macromolecule interactions govern the formation of lubricating films, through both adsorption and the formation of thicker, aggregated films. Adsorbed film properties were strongly correlated with the lubricating film thickness for protein solutions, but affected by structural changes for solutions containing lipid vesicles. Comparing the lubrication of current and cutting-edge materials shows that the permeability of the surface influences the active lubrication mechanisms. Based on the findings of this work, recommendations for the composition of model synovial fluids are made.

The main objective of this work was to identify the effect parameters in the method of material testing have on three constitutive models that are commonly used to predict high strain rate, large strain material behaviour. These are the Goldthorpe path-dependent model, the Goldthorpe-modified Armstrong-Zerilli model and the path-dependent fracture model, also developed by Goldthorpe. To investigate the material models, numerous interrupted tensile, torsion and compression tests were performed. This led to key improvements to the image analysis and acquisition steps of the interrupted tensile test methodology used to quantify necking. Whilst the models used to predict dynamic deformation and damage to very high strains for ballistic impact applications, such as the Goldthorpe path-dependent failure model, incorporate temperature and strain-rate dependence, they do not consider the effect of specimen size or the microstructural length scale. To investigate this, geometrically similar specimens spanning a scale range of 100:1 were tested quasi-statically to failure. Images of neck evolution were acquired using optical techniques for large specimens, and in-situ scanning electron microscope testing for small specimens, to examine the dependence of neck geometry on a broad range of specimen sizes. Size effects typically arise when the smallest specimen dimension is on the order of a microstructural length (e.g. grain size, dislocation mean free path, etc.), or in the presence of significant plastic strain gradients, which increase the density of geometrically necessary dislocations. An assumption also made to develop the models is that material deformation is isotropic. This encouraged a study where calibrated cameras, multiple view geometry and edge detection tools were used to measure the anisotropic deformation of uniaxially loaded cylindrical specimens. For this investigation tension tests were performed at quasi-static strain rates on a variety of materials using screw-driven test machines. The elliptically evolving specimens were then measured at multiple interruptions to determine the true direct stress and strain at the neck. It is hoped the anisotropic data presented in this thesis will aid model development.

Creep is the primary life limiting mechanism of static high temperature, high pressure power station components. Creep state evaluation is currently achieved by surface inspection of microstructure during infrequent outages; a methodology which is laborious, time consuming and considered inadequate. The objective of this work is to develop a monitoring technique that is capable of on-load creep damage monitoring. A continuous update of component integrity will enable better informed, targeted inspections and outage maintenance providing increased power generation availability. A low-frequency, permanently installed potential drop system has been previously developed and will be the focus of this thesis. The use of a quasi-DC inspection frequency suppresses the influence of the electromagnetic skin effect that would otherwise undermine the stability of the measurement in the ferromagnetic materials of interest; the use of even low frequency measurements allows phase sensitive detection and greatly enhanced noise performance. By permanently installing the electrodes to the surface of the component the resistance measurement is sensitive to strain. A resistance - strain inversion is derived and validated experimentally; the use of the potential drop sensor as a robust, high temperature strain gauge is therefore demonstrated. The strain rate of a component is known to be an expression of the creep state of the component. This concept was adopted to develop an interpretive framework for inferring the creep state of a component. It is possible to monitor the accumulation of creep damage through the symptomatic relative increase in strain rate. By taking the ratio of two orthogonal strain measurements, instability and drift common to both measurements can be effectively eliminated; an important attribute considering the necessity to monitor very low strain rates over decades in time in a harsh environment. A preliminary study of using the potential drop technique for monitoring creep damage at a weld has been conducted. Welds provide a site for preferential creep damage accumulation and therefore will frequently be the life limiting feature of power station components. The potential drop technique will be sensitive to both the localised strain that is understood to act as precursor to creep damage at a weld and also the initiation and growth of a crack. Through the course of this project, two site trials have been conducted in power stations. A measurement system and high temperature hardware that is suitable for the power station environment has been developed. The focus of this thesis is the effective transfer of the technique to industry; the realisation of this is detailed in the final chapter.