An attempt has been made to study bond strength history in Prestressed Concrete Reactor Vessels (PCRV) which house the Advanced Gas-cooled Reactors. Three-dimensional non-linear analytical model has been developed in which the effect of bond is included. A finite element computer program is written in which solid, membrane, line and bond-linkage elements have been used to represent vessel concrete, steel liner, pre-stressing tendons and bond (between steel and concrete) respectively. Concrete is assumed to be non-linear material in compression and linear brittle (tension cut-off) material in tension, and the steel as elasto-plastic material with strain hardening. Provision is also made for concrete cracking, crushing and visco-elastic creep. Two experiments have been carried out during this research. The purpose of the first experiment was to determine bond coefficients required for the analysis. This was achieved by pull-out tests on prestressing specimens using 5 mm and 7 mm diameter prestressing wires. The second experiment was performed on an octagonal prestressed concrete slab representing a top cap of a reactor vessel. The experimental results obtained from this slab are corroborated with the analytical results. A typical Prestressed Concrete Reactor Vessel with boilers and circulators housed within the vessel wall thickness has been analysed for bond strength under increasing gas pressure at suitable intervals of its 30 year life. A comparison is made between the unbonded and bonded vessels. All analytical results compare well with those obtained from the experiments and available published data.

The safety of infrastructure facilities is the primary objective of any civil engineering design. A large section of underground pipelines in the UK are classified as structurally deficient and functionally obsolete. Due to low visibility, condition assessment and rehabilitation of underground pipelines are frequently neglected until a catastrophic failure occurs. Providing an acceptable level of service and overcoming the practical difficulties, the concerned industry has to plan how to operate, maintain and renew (repair or replace) the pipeline systems under the budget constraints. This research is focused on estimating pipe reliability and deciding when and how interventions are needed to prevent unexpected failures of flexible underground metal pipelines subject to externally applied loadings and pipe material corrosion during the whole service life at the optimal cost. The time-dependent reliability due to corrosion induced excessive deflection, buckling, wall thrust and bending has been estimated. First, Hassofer-Lind and Rackwitz-Fiessler (HL-RF) algorithm and Monte Carlo Simulation (MCS) have been used to estimate the reliability. Then Subset Simulation (SS) method is developed to enhance the applicability, especially for small failure probability prediction. Accuracy prediction method, Receiver Operating Characteristic (ROC) curve has been introduced in this research to assess the accuracy of pipeline reliability analysis. Then the study has been extended to determine the intervention year for maintenance and identify the most appropriate renewal solution by optimising the risk of failure and life cycle cost, including carbon dioxide emissions mitigation cost, using Genetic Algorithm (GA). Optimisation technique, SS has also been developed for risk-cost optimisation of underground pipelines. Examples are presented to validate the proposed methods with a view to prevent unexpected failure of pipes by prioritising maintenance based on failure severity and system reliability. The proposed reliability estimation and risk-cost optimisation approach can be utilised to form a maintenance strategy and to avoid unexpected failure of pipeline networks during service life.

This thesis is concerned with a research programme into a triangulation based Surface Topography Measurement (STM) system. The particular system investigated produces a series of three dimensional coordinates of points on a target surface which has been placed in front of the STM system. It is designed for use at long ranges (5m - 20m initially) and onto rough surfaces (a selection of natural materials with particle sizes up to 1cm). The system uses a laser triangulation method with a single spot projected onto the target surface and through movement of the spot over the surface a range map is built up. A simple model which describes the STM system is presented along with an equation which gives the range error, any effects on the model by the physical implementation of the STM system are then described and quantified. Performance tests carried out on the system are described and a second version of the system is then outlined which addresses deficiencies isolated in the original system. A more detailed discussion of how the range errors arise is also given. New hardware enables the received laser spot's distribution to be analysed and processed so that, in certain cases, subpixel resolution of the centre of the laser spot can be obtained. Expansion of the laser spot and varying the parameters of the optical system are examined to assess their effect over different target materials and estimates of the possible range accuracy are also given. These results presented will be used to assess the applicability of the system for use in an industrial environment. Further work connected to this research area and several industrial applications are also detailed.

Input Data Management (IDM) is a time consuming and costly process for Discrete Event Simulation (DES) projects. Input Data Management is considered as the basis of real-time process simulation (Bergmann, Stelzer and Strassburger, 2011). According to Bengtsson et al. (2009), data input phase constitutes on the average about 31% of the time of an entire simulation project. Moreover, the lack of interoperability between manufacturing applications and simulation software leads to a high cost to users in developing custom-built interfaces (NIST, 2011). Therefore, a standard structure is needed to facilitate data exchange within different manufacturing data from various sources and simulation packages (Harari, 2012). This work aims to develop an IDM system in order to enhance automatic data exchange within manufacturing process simulation projects. The approach is to use a Business Intelligence (BI) package to extract and collect simulation data from various sources, and then import to a DES application. Thus, a simulation model can be automatically updated in DES application by real-time system data. The prototype system is currently implemented in pilot-scale, and results are based on sample data files retrieved from the Simulation Engineering Department at Ford Motor Company. In addition to automatic data management, the system also generates an interoperable data file, which is compatible with most DES packages for simulation experiments. For the first time, a Business Intelligence system is applied to simulation data management projects, and is introduced as a novel solution for this context. The results of pilot-implementation indicated the possibility of direct connection between raw data sources and model database. The direct connection allows DES applications to integrate with different business systems without the need of plug-in translators for interpreting the exchanged data file. The system developed also reduces the number of human-driven processes through model update procedures. Also, it is found that the BI package can be integrated with existing middleware used in a company, to maintain an “as-is” simulation procedure, and minimise the possibility of process change in systematic model update.

The granular material behaviour is determined by the local contact behaviour between particles and the spatial arrangement of particles. Investigation of particle-scale mechanism provides fundamental insights into global granular material behaviour. A multi-scale investigation has been carried out to study granular material behaviour under general stress paths using discrete element method (DEM). The commercial software Particle Flow Code in Three Dimensions (PFC3D) is employed for numerical simulations and the linear contact model is used to describe local contact behaviour. General loading paths were achieved by implementing a boundary control programme with independent control of both the magnitudes of three principal stresses and their principal directions. The intermediate principal stress ratio , where are the major, intermediate and minor principal stresses, and material anisotropy both had significant effect on granular material strength. The true triaxial simulation results indicated that the peak stress ratio was mainly contributed by the micro-scale contact force anisotropy. A smaller stress ratio was observed at greater a b value due to smaller degree of contact force anisotropy. Fabric anisotropy was another contributor to the material stress state. A lower peak stress ratio was obtained at a larger tilting major principal stress direction from the vertical deposition direction since smaller fabric anisotropy degree developed at larger . However, the material initial anisotropy had negligible effect on the critical stress ratio owing to the same contact force anisotropy and fabric anisotropy achieved. In true triaxial simulations, the intermediate strain increment rate ratio was generally larger than the stress ratio b since the particle-scale tangential force ratio was observed to be smaller than b value. The non-coaxial deformation observed in monotonic loading with various loading direction can be explained due to the non-coincidence between the principal fabric direction and the principal stress direction. And the degree of non-coaxiality decreased against shearing as the principal fabric direction approached loading direction gradually. The granular material response to rotational shear showed significant volumetric contraction and deformation non-coaxiality. The material internal structure rotated continuously along the principal stress rotation. The principal fabric direction did not exactly follow the rotation of principal stress direction. The fabric reorganisation mechanism accompanied by irrecoverable plastic deformation, leading to non-coaxial deformation behaviour. During rotational shear, the ultimate void ratio was determined by the stress ratio and b value but independent of initial void ratios. Under otherwise identical conditions, the greater internal structure anisotropy was observed at the higher stress ratio and at a greater b value, resulting in smaller ultimate void ratio (larger volumetric contraction). The general degree of deformation non-coaxiality decreased with increasing stress ratio and b value for rotational shear. The difference between the major principal stress direction and the major principal fabric direction was smaller at higher stress ratio and greater b value. It was interesting to note that the sample could fail during rotational shear, resulting in significant deviatoric strain developed in the first few cycles. The sample failed at a stress ratio , which was lower than the peak stress ratio obtained in monotonic loading but higher than the critical stress ratio . This indicated importance of considering stress rotation in geotechnical design and the material strength should be chosen based on the critical stress ratio rather than the peak value. The multi-scale investigation of granular material explains the strength characteristics from the micromechanical point of view. Observations on the fabric evolution have been made under various loading conditions. This may be useful information for the development of an advanced constitutive model.

There has been a growing interest in using discontinuous carbon fibre composites for semi-structural applications within the automotive industry. The main advantages of discontinuous fibres are low material costs, low wastage and low touch labour compared with processes using carbon fibre textiles. Directed Carbon Fibre Preforming (DCFP) is an automated process for producing complex 3D preforms for liquid moulding. DCFP offers the potential for producing highly optimised structures, with local control over tow size, fibre length and volume fraction within the component. The execution of this is challenging however, as confidence in the current library of material properties is low and existing structural optimisation packages only consider a very limited number of design variables, which are restricted to more conventional composite materials. This thesis aims to establish a structural design tool to exploit the design freedom offered by the DCFP process. A large number of parameters associated with the fibre architecture can be controlled to meet a range of design criterions such as performance, weight and cost. The optimisation tool is capable of generating locally varied fibre areal mass and thickness maps that are suitable for manufacture by the robot controlled process. The developed model adopts a multi-scaled finite element approach. Meso-scale simulations are performed to establish size effects in discontinuous fibre composites, to quantify the level of stochastic variability and to determine the representative volume element for a given fibre architecture. A DCFP material database is generated to facilitate macro-scale modelling at the component level. The macro-scale model iteratively redistributes material in order to minimise the total strain energy of the model under prescribed loading conditions. The optimised model is segmented into areas of uniform areal mass, where the zone geometries are tailored to achieve representative material properties according to the meso-scale results, whilst ensuring the design is fit for manufacture. An automotive spare wheel well has been chosen as a demonstrator component, enabling two DCFP architectures to be compared against a continuous glass/carbon fibre NCF design. The first case offers a high performance (high specific stiffness) solution and the second offers a low cost option using high filament count tows. Following optimisation, results suggest that a 3K 25mm fibre length DCFP option can achieve a specific stiffness 52% higher than the glass/carbon baseline design, but for 1.33 times higher material cost. Alternatively, the specific stiffness of a 24K 50mm fibre length DCFP is marginally lower than the first option, but still out-performs the baseline for just 67% of the material cost. The structural optimisation method demonstrates that discontinuous fibre composites can compete against continuous fibre counterparts for semi-structural applications.

The Scale Invariant Feature Transform (SIFT) is one of the most popular matching algorithms in the field of computer vision. It takes over many other algorithms because features detected are fully invariant to image scaling and rotation, and are also shown to be robust to changes in 3D viewpoint, addition of noise, changes in illumination and a sustainable range of affine distortion. However, the computational complexity is high, which prevents it from achieving real-time. The aim of this project, therefore, is to develop a high-performance image matching system based on the optimised SIFT algorithm to perform real-time feature detection, description and matching. This thesis presents the stages of the development of the system. To reduce the computational complexity, an alternative to the grid layout of standard SIFT is proposed, which is termed as SRI-DASIY (Scale and Rotation Invariant DAISY). The SRI-DAISY achieves comparable performance with the standard SIFT descriptor, but is more efficient to be implemented using hardware, in terms of both computational complexity and memory usage. The design takes only 7.57 µs to generate a descriptor with a system frequency of 100 MHz, which is equivalent to approximately 132,100 descriptors per second and is of the highest throughput when compared with existing designs. Besides, a novel keypoint matching strategy is also presented in this thesis, which achieves higher precision than the widely applied distance ratio based matching and is computationally more efficient. All phases of the SIFT algorithm have been investigated, including feature detection, descriptor generation and descriptor matching. The characterisation of each individual part of the design is carried out and compared with the software simulation results. A fully stand-alone image matching system has been developed that consists of a CMOS camera front-end for image capture, a SIFT processing core embedded in a Field Programmable Logic Array (FPGA) device, and a USB back-end for data transfer. Experiments are conducted by using real-world images to verify the system performance. The system has been tested by integrating into two practical applications. The resulting image matching system eliminates the bottlenecks that limit the overall throughput of the system, and hence allowing the system to process images in real-time without interruption. The design can be modified to adapt to the applications processing images with higher resolution and is still able to achieve real-time.

Having knowledge of the current creep strength of service-aged components in high temperature installations such as nuclear power stations, oil refineries and chemical plants is essential for their safe and economic operation. Obtaining this knowledge may involve the use of small material samples. These small samples may be removed from weld regions or from component surfaces. Improving small specimens creep testing techniques, whereby a reliable uniaxial minimum strain rate and rupture data can be obtained, has been a major engineering concern for the last 20 years or so. This thesis includes the development of the small ring creep testing specimen in order to allow the ring specimen to be manufactured and tested with various shapes and geometries. The shape and size of the available small material samples normally dictates the ring shape, e.g., circular or elliptical. However, changing the ring shape leads to a change in the conversion factors, which are used to convert the ring data to the corresponding uniaxial data. Therefore, the effects of the ring geometry with different thicknesses, on the conversion factors, are described in this work. The finite element analyses have been used to assess the effects of shear deformation on the ring behaviour and also to determine the optimum ring ii geometry. Nickel base Superalloy 738 steel at 800oC and (Bar-257) P91 steel at 650oC have been used to validate the testing method. Two new small sized creep test specimens are also described in this thesis, i.e., (i) a small (Two-bar) specimen, which is suitable for use in obtaining the uniaxial MSR and creep rupture data and (ii) a small notched specimen which is suitable for obtaining the multiaxial stress state parameter. The specimen testing techniques, modeling, loading and manufacturing are described for both specimen types in this work. Finite element analyses have been used to assess the effects of the two-bar specimen (TBS) dimensions on the conversion factors, the failure time, the minimum strain rate, and to determine optimum dimension ratio ranges for the specimen. The two-bar specimen and the small notched specimen have been used to obtain a full set of material constants for two high temperature materials, i.e., (i) typical (as received) P91steel at 600oC and (ii) weak (Bar-257) P91 steel at 650oC. The results show remarkably good agreement between the data obtained from the two new small specimen testing techniques and the data obtained from corresponding uniaxial tests. The major advantages of the small ring specimen, the two-bar specimen and the small notched specimen testing techniques, over the existing small specimens testing techniques, are also included.

Polymer-based nanocomposites are of significant current research interest owing to their outstanding mechanical properties, light weight, processability and low cost. They are also increasingly being considered for a range of industrial applications, including packaging, fuel tanks, gas barriers and high performance films. Ultra-high molecular weight polyethylene (UHMWPE) is already used in various applications, such as lightweight body armour because of its high impact resistance with light weight and total joint replacement due to its high wear resistance. However, a broader use of UHMWPE is limited by the complexity and cost of the manufacturing process, which can be attributed to its high viscosity at processing temperatures. The processability of UHMWPE can be improved by blending with a compatible, lower molecular weight polymer, however, this inevitably results in a reduction in some of the useful properties, such as impact resistance. In this work the potential of adding nano-fillers to such blends to create a range of nanocomposite polymers with the advantages of easy processability and enhanced properties is investigated. The overall aim of this research was to investigate the effect of processing method, strain rate, nanoparticle type and content on the morphological, thermal and mechanical properties of a family of novel polyethylene-based nanocomposites. Polymer nanocomposites of blended UHMWPE and high density polyethylene (HDPE) reinforced with carbon black (CB), carbon nanotubes (CNTs) or inorganic clay were prepared using conventional processing techniques. After initial experiments into the effects of processing parameters, two sets of processing parameters were selected that gave different blend morphology in order to investigate the effect of this on the blend properties and nanofiller dispersion. Characterization of the pure, blended and nanocomposite materials was achieved by the application of combination of experimental techniques. Tensile testing was carried out to characterise the effect of processing method, strain rate, ambient temperature, nanoparticle type and content on the stress-strain behaviour and also to study heat generation during plastic deformation at high strain rates. Depth sensing indentation (DSI) tests were carried out to characterise the effect of processing method, ambient temperature, nanoparticle type and content on the near-surface properties of the materials at a micro-scale under a more complex state of stress that more closely approximates that seen in impact applications. The creep behaviour of the materials was investigated at macro and micro scales at various ambient temperatures. This is important as a weakness of UHMWPE is poor creep resistance and it would be extremely useful if blending or the addition of nanofillers could improve this. A phenomenological model was used to analyse the creep data as this can be usefully used to predict creep performance in service and to aid understanding of the creep phenomena in these materials. The results included in this work are summarised below. Firstly, it was seen that processing parameters had a significant effect on the morphology of the blends, which in turn affected the blend properties and the dispersion of nanoparticles in the blend. Secondly, it was seen that heat generation during plastic deformation of the polyethylene blends and nanocomposites was significantly dependent on morphology, strain rate, nanoparticle type and content. Furthermore, this temperature increase strongly affected the material properties at high strain rates, which is an important consideration if these materials are to be used in high strain rate applications, e.g. as replacement for UHMWPE in helmets and body armour. Thirdly, the macro and micro viscoelastic behaviour of the materials was strongly dependent on the morphology, nanoparticle type and content. A significant increase in creep resistance compared with UHMWPE could be engineered by a careful selection of blend and nanoparticle type and weight fraction. It can be seen, therefore, that a new class of cheap and easy processable polymer nanocomposites have been characterised that can give a range of property sets dependent on the blend processing and nanofiller type and weight fraction. Although certain compromises in property sets are unavoidable, e.g. it is difficult to engineer maximum creep and impact resistance in the same material, this ability to tailor properties could potentially increase the range of applications for these materials and enable better product design.

This thesis focuses on flow through compliant porous media with applications to the manufacturing of composites by vacuum infusion (VI). The context of this work is the need for reliability in environmentally friendly composite processing methods for composite materials. Commercial reality and the prospective application to low cost structures for the transportation industry dictate that appropriate emphasis should be put on obtaining robust simulations, ensuring reliability and progressing toward efficient means of process control. In this context, the open mould manufacturing processes which have been used to produce large composite structures, and are not conducive to quality nor environmental responsibility, must be replaced. Hence, establishing composites as a viable alternative requires closed moulding techniques, of which VI is the most practical for large structures, but where reliability is required for economic survival. This work addresses many aspects of this problem, by making innovative use of fluid mechanics and developing, implementing and proposing new analysis and modelling tools for VI. Main results include a validated analytical model for flow through compliant media, a study of the compliance of textile reinforcements, a finite element model for VI and novel stochastic techniques for the analysis of reliability in liquid composite moulding processes. The work discussed herein stems from a thorough evaluation of published models and leads to novel flow modelling tools for VI including a unique and general formalism for textile compliance. Using these tools it was possible to study, for the first time, the effect of different parameters on VI manufacturing. The reliability issue was addressed by integrating stochastic models for compliance and permeability, and the ability to model complex geometries was demonstrated by adapting a commercial finite element flow code (LIMS).