This thesis includes an experimental investigation of the behaviour of steel fibre reinforced high performance concrete (SFRHPC) under biaxial loading conditions. Also included are constitutive models to enable numerical predictions of the strength behaviour of such a material. Within the experimental stage a large biaxial test machine was designed and manufactured. The load capacity of each axis was 2000 kN. Special thought was given to the load platen system because of the friction which occurs between the platen system and the concrete specimen. Brush bearing platens and solid steel blocks with and without Teflon friction reducing pads were tried. Because the brush platen and the Teflon pads were constantly damaged during testing, solid steel block platens were finally used. For tests where tension loads were involved, experiments were carried out with dog-bone shaped specimens and specimens glued on to the platens. Finally, the tension loads were transferred through threaded rods cast into the specimens and connected to the machine patterns with screws. Modern control schemes and high speed data acquisition systems were used to monitor the material response and to collect experimental stress and strain results. The principal deformations were monitored and the crack patterns and failure modes examined. Failure envelopes were developed based on the strength data for each fibre variable. The load capacity of SFRHPC under biaxial load conditions was found to be larger than for plain HPC for all fibre types and volume fractions. The stress-strain recordings indicated a linear behaviour almost up to failure. The examined failure mode between plain and steel fibre HPC was similar and all specimens failed very suddenly with a splitting failure type. The test variables included four different types of hooked ended steel fibres with different fibre volume fractions from plain to 2% in 0.5% steps. The specimens were tested under the entire range of stress combinations including uniaxial compression and tension and biaxial compression-compression and compression-tension. As a result the specimen still failed in an explosive manner but the pieces were still connected together by fibres after failure. The biaxial strength compared to the uniaxial strength increased more, as observed with normal strength concrete. With the addition of fibres the biaxial strength behaviour of HPC was almost the same as for plain normal strength concrete. The deformation characteristics of plain and SFRHPC showed a linear behaviour up to a higher stress than normal strength concrete. In fact the linear limit was almost as high as the failure load. The examined strength data was used to model the biaxial strength envelopes of HPC and SFRHPC using different methods. These included the Ottosen failure criterion and the Willam and Warnke failure criterion.

The main aim of the presented work is the development of a reliable and coherent solution approach to investigate thermo-hygro-mechanical behaviour of concrete, especially under severe heating conditions. The work focuses on the development and extension of an existing analytical, numerical and constitutive model developed at the University of Glasgow. This is then used as a predictive modelling tool to investigate the response of concrete structures subject to combined thermo-mechanical loads. The thesis focuses initially on the coupled heat and mass transport model. A novel alternative formulation for sorption isotherms, that is applicable to both normal and high strength concrete, is developed. Furthermore, the effect of the slip flow effect is included and several state equations, including relative permeability and saturation vapour pressure, are adopted. Additionally, the effect of polypropylene fibres as a spalling prevention technique is modelled via a modification of the intrinsic permeability formulation. The transport model is further coupled with a damage-based mechanical model for concrete. The fully coupled thermo-hygro-mechanical model is presented through validation and verification problems and case studies. The model is implemented in a finite element formulation and behaves in a robust manner. The predictions of moisture state for the benchmark problems of drying and heating compare well with experimental results. Classical behaviour associated with heated concrete, such as moisture clog, gas pressure build-up, etc. are all captured by the presented model. The thesis concludes by considering the analysis of prestressed concrete pressure vessel.

This thesis presents a study of punching shear capacity of flat slab-column junctions. A three dimensional non-linear finite element program based on 20 node isoparametric solid element was used for the investigation. The non-linear 3-D elastic isotropic model proposed by Motsovos was used to describe the behaviour of concrete before cracking or crushing. After cracking, a yield criteria for 2-D model similar to Kupfer-Hilsdorf was used and concrete was assumed to be anisotropic. No softening in compression is assumed. Smeared crack approach with simple tension stiffening and shear retention equations were employed to mimic the post-cracking behaviour of concrete. Reinforcing bars were represented by one dimensional element embedded in the solid elements and for both tension and compression, linear elastic-plastic behaviour is assumed. A comparison was first made between the predictions of slab behaviour using Kotsovos' model (In-house program) and plasticity based model used in the commercial package ABAQUS. From this it was concluded that Kotsovos' concrete model is a good model for the three dimensional analysis of the punching shear problem but the ABAQUS model was unsuitable. In order to achieve an accurate and economical solution for the non-linear analysis, a parametric study was carried out to choose a suitable analytical model. After having chosen the "best" concrete model, over 175 slabs from different sources were analysed using a constant set parameters. The analysis includes various types connections (interior, edge and corner) with and without shear reinforcement, subjected to shear force alone or to a combination of shear force and unbalanced moment. These slabs cover most of the factors affecting punching shear strength, such as slab thickness, flexural reinforcement ratios, concrete strength, and column size. This study also includes the effect of in-plane restraint on punching shear strength of slabs.

The object of this Ph.D work is to evaluate fidelity in surface measurements in nanometric metrology for both contact and non-contact methods, namely stylus instruments and scanning tunnelling microscopy. Fidelity is defined, in this thesis, as a measure to which an instrument system reproduces the surface features and thus the parameters of interest. High fidelity measurement has two meanings; less distortion in the measured result and less disturbance to the surface being measured. Interaction at the interface between the probe and the surface is the source of failure to achieve high fidelity. No instrument measures surface topography alone: all instruments measure a convolution of topography and the geometrical and physical interaction of the measured probe and the surface. In the case of a mechanical stylus, factors extraneous to the topography include (a) the shape and size of the stylus, (b) mechanical properties of the stylus and the specimen such as elastic moduli and hardness, (c) frictional force of the sliding pair. and (d) dynamic interaction forces during the sliding. For the scanning tunnelling microscope, factors which affect measurement in addition to topography include the geometry of the tip, the electronic properties of the surface and mechanical deformation due to electrostatic forces and contamination. 'These factors have been investigated in great detail, particularly for the stylus instruments. A specially designed electro-magnetic force actuator has been developed to give a better control on loading force during the experiments. Tracking force effects were evaluated by profiling statistical parameters, and scanning electron microscopy. Friction between a stylus and specimen has been measured for different loading force, sliding speed, material and surface finish. Improvement on dynamic characteristics of a stylus system has been achieved by active damping control. An optimal damping ratio for stylus instruments is found to be within 0.5-0.7. Through the study, the tracking force and traversing speed are found to be the crucial factors to be tackled so that high fidelity measurement can be obtained. A similar investigation has been also made on two home-built scanning tunnelling microscopes to explore the potential applications of STM on nanometric metrology.

Reconstructing depth information from images is one of the actively researched themes in computer vision and its application involves most vision research areas from object recognition to realistic visualisation. Amongst other useful vision-based reconstruction techniques, this thesis extensively investigates the visual hull (VH) concept for volume approximation and its robust surface modelling when various views of an object are available. Assuming that multiple images are captured from a circular motion, projection matrices are generally parameterised in terms of a rotation angle from a reference position in order to facilitate the multi-camera calibration. However, this assumption is often violated in practice, i.e., a pure rotation in a planar motion with accurate rotation angle is hardly realisable. To address this problem, at first, this thesis proposes a calibration method associated with the approximate circular motion. With these modified projection matrices, a resulting VH is represented by a hierarchical tree structure of voxels from which surfaces are extracted by the Marching cubes (MC) algorithm. However, the surfaces may have unexpected artefacts caused by a coarser volume reconstruction, the topological ambiguity of the MC algorithm, and imperfect image processing or calibration result. To avoid this sensitivity, this thesis proposes a robust surface construction algorithm which initially classifies local convex regions from imperfect MC vertices and then aggregates local surfaces constructed by the 3D convex hull algorithm. Furthermore, this thesis also explores the use of wide baseline images to refine a coarse VH using an affine invariant region descriptor. This improves the quality of VH when a small number of initial views is given. In conclusion, the proposed methods achieve a 3D model with enhanced accuracy. Also, robust surface modelling is retained when silhouette images are degraded by practical noise.

The main research theme of this engineering doctorate was as follows: 'How can UK industry effectively leverage concurrent engineering practices within the organisation?' To address this problem, a review of implementation tools was undertaken, where a number of 'do it yourself' methodologies were identified within the literature for implementing Concurrent Engineering (CE). These were 'change management workbooks' and 'self-assessment tools'. Change management workbooks provide a means for managing an implementation program, and self-assessment measures and benchmarks the deployment of practices and identifying areas for improvement. However, both were found to operate independently from one another. Therefore, it has been proposed that the deployment of CE practice can be further enhanced through an integrated approach. This approach combines both disciplines, because as a system it could measure the deployment of practices, identify future improvements, and enable an organisation to manage the transition to better CE. To solve this problem a self-assessment tool, which encapsulated practice from the automobile, power generation, aerospace, pneumatics and mechatronics industry sectors was developed. The tool consisted of both a practice and performance dimension to ensure that practices deployed impacted the bottom line. The selfassessment tool allows for a company to assess its current state, and decide where it wants to go. Furthermore, a change management process was developed, which integrated with the self-assessment tool, and a series of additional tools developed specifically for the task such as the generic planning tool, a decision tool for deciding an implementation strategy, and additional tools, which have been selected from the literature. The application of the system at an UK based organisation demonstrated that a self-assessment tool integrated with a change management process can assist a facilitator to direct a change program toward implementing CE practice. Furthermore, the preliminary results of this application illustrate that the practices deployed as a result of the system were impacting key project performance measures.

Presented in this thesis is an evaluation of a novel box beam system of Pul- truded Fibre Reinforced Polymer (PFRP) shapes. The flat-pack modular beam system consists of separate PFRP flange and web shapes joined together with a new method of mechanical fastening. It is based on the first generation Star- tlink building system, conceived by UK engineers in 1999. The Startlink building system is introduced, and classified within the scope of Modern Methods of Con- struction (MMC), and its merits are discussed. In the context of MMC a critical review by the author finds that, although the proposed 1999 generation Startlink system offers design flexibility, it will probably have a limited market potential. The novel use of the steel MlO Unistrut connection method as a means of fastening distinct PFRP shapes in a building system is characterised. Individ- ual connector design parameters for joint stiffness and resistance are identified and determined, under pure shear loading. The results of a series of physical tests show no significant loss of stiffness or strength with long term environmen- tal exposure. Values of key mechanical properties for design calculations are recommended. A 400 x 200 x 2848 mm prototype PFRP box beam assembly is fabricated from two flange and two web panel-type shapes, cut from existing off-the-shelf PFRP shapes. This is 60 mm deeper than the largest single PFRP shape that could be used as a beam. The assembly is joined at the web-flange junction with M10 Unistrut connectors set at various spacing's, in the range 50 to 400 mm. These connectors carry the longitudinal shear that is generated between the joined shapes when the modular assembly is in flexure, Theoretical deflections, cal- culated using a modified form of a partial-interaction analysis model developed for composite concrete and steel structures, are predicted for the assembly ac- counting for the finite shear stiffness of the web-flange connection. A series of 16 four-point bending load tests on the beam assembly, across two load arrange- ments, show that its performance is linked to the designated spacing of the M10 Unistrut connections. The flexural rigidity and degree of interaction present in the assembly are determined from analysis of vertical deflections and longitudi- nal strains, as the beam is deformed. The influence of secondary effects, due to the poor tolerances achieved in the hand fabrication of the beam's assembly, are found to greatly affect the ability of the deflection analysis to give the required measured deflections. Comparison of the effective joint shear rigidities obtained from theory and testing indicates a higher individual connection stiffness in the prototype beam than previously determined by way of the individual Unistrut connector characterisation. It is found through the combined analytical and physical testing research that the M10 Unistrut connection method can only provide the necessary joint shear stiffness and resistance to the 400 mm deep beam if the connector spacing, along the four joints, is ≤ 50 mm. The total number of connectors this represents in the beam is likely to make this modular construction approach too expensive for it to be commercially viable. Although the M10 connector could be used to fabricate beams of lesser depths, since the number of connectors will then be reduced, these beams would find it difficult to compete with the available off-the-shelf PFRP beam shapes, of up to 300 mm deep. There is however scope to use the Unistrut method of connection to provide longitudinal shear resistance in building systems where, for example, a floor panel is to be stiffened by a channel shaped beam and the overall depth is ≤ 300 mm. The research work contained in this thesis has contributed to a radical change in the PFRP product offerings now proposed in the 2006 generation Startlink building system.

In my portfolio I have attempted, through projects and post module work in both a large and a small start up company, to examine consulting engineering firms as businesses. Firstly identifying the key issues then developing a framework of ideas that could be used by others. Because of the lack of literature directly related to the practice of consultipg engineering in construction I have looked outside that industry for references to guide me. In this summary I have attempted to consolidate the information gained. This would be useful as a starting point when assessing an existing businesses or as guidance for start-up firms.

In this thesis, two problems have been studied: one is the reliability of design resistances of composite beams in bending, the other is the longitudinal shear resistances in concrete flanges of composite beams with profiled sheeting. For the first problem, a general theory is developed based on the concepts of safety-index and second-moment reliability, to analyse the reliability of a design resistance for composite structures. This theory consists of two basic parts: one is the determination of several distinct safety factors for a single resistance function, with respect to the specified safety level; the other is the assessment of safety level, corresponding to given safety factors. Application of this theory is related to the relevant test results. In the scope of Eurocode 4, this theory is particularly used to calibrate safety factors employed for design plastic bending resistances of composite beams. From the calculation based on 122 test results, the reliability of using the Eurocode provisions, including the recommended safety factors and the theoretical models, is judged for bending resistances of beams with full and partial shear connection. In studying the longitudinal shear resistances in concrete flanges of composite beams with and without profiled sheeting, a simple theory to determine the resistances is established. This theory is based on the "friction-cohesion" and "concrete strut" models for concrete in shear and takes the effect of profiled sheeting into account. Associated with the theoretical work, seven tests on three specimens for investigating the longitudinal shear resistances are reported. From both experimental results and theoretical calculation, the behaviour of the test specimens is studied, and, finally, new design recommendations are made to predict the longitudinal shear resistances of the composite beams.

Industrial globalisation and the increased complexity of products promote new product development (NPD) by distributed teams. These teams facilitate the transfer of designs to manufacturing facilities, the adaptation of products to local markets and the access to engineering design talents. However, there are many industrial experiences of under performing distributed teams, such as the development of the Ford Mondeo which ran a year late and cost £4 billion. This Engineering Doctorate project CETGI (Concurrent Engineering and Teamwork across Global Industries) aimed to investigate and advise companies on the enabling practices of global concurrent engineering (GCE). An assessment tool for GCE was developed, embedding its enabling practices in a knowledge base and providing structured advice to manufacturing companies of electro-mechanical products. Preliminary investigation was conducted towards the development of the CETGI tool: literature review on GCE; a benchmark study of nine current concurrent engineering assessment tools; a comparative analysis of NPD practices in the UK and the USA; a survey of NPD practices of global companies developing products in the UK; and thirteen case studies of GCE projects. The CETGI tool was then developed, consisting of an assessment process involving multifunctional teams and executives. CETGI is different to previous tools because of its knowledge base of GCE practices and its three analytical algorithms: the first, aggregates the individual answers; the second, provides a customised list of prioritised actions; and the third, generates maturity charts, providing a performance perspective and supporting benchmarking. Another novel feature of CETGI is the Internet enabled software application that supports its assessment process. The latter software is a multi-participant decision-making support system. The CETGI tool has been applied to three first tier supplier companies, two from the automotive industry and one from the transport industry aimed at fine-tuning and validating the tool. The work focused on evaluating the reliability of the questionnaire (constructed on the knowledge base) and establishing the validity and the industrial usefulness of the CETGI tool. On average, the NPD executives considered 90% of the suggested actions appropriate, selected 70% of the actions for implementation and implemented 51% of the selected actions (based on a post-assessment review at one company). The actions were estimated to reduce time-to-market and product-unit costs by the NPD executives and they rated CETGI highest in terms of promoting global product development, followed by providing an overall picture of the current product development process. Further research opportunities have been identified such as using the CETGI tool to assess other business areas apart from GCE. This would require investigating best practices in these areas and embedding the results in specific knowledge bases. The assessment method, the structure of the knowledge base and the analytical algorithms would remain the same.