Corrosion of reinforcing steel is a major serviceability issue with reinforced concrete structures, often resulting in significant section and bond loss. However, current non-destructive diagnostic techniques do not allow corrosion to be reliably detected at the very early stages of the process, before damage to the concrete occurs. This research describes the development of an Acoustic Emission (AE) technique as a practical tool for the early detection of corrosion of reinforcing steel embedded in concrete. The study falls into three main areas: (i) determining the influential material parameters of reinforced concrete that affect the magnitude of the acoustic emissions; (ii) investigating the influence of diurnal and seasonal temperature variations on corrosion rate and thus the rate of acoustic emissions; and (iii) developing a testing and evaluation procedure that combines the findings of the first two stages with existing knowledge about corrosion and deterioration of concrete structures. In the first phase of the research material parameters such as cover thickness, compressive strength and rebar diameter were investigated to ascertain the influence of varying these factors on the magnitude of AE Energy obtained per gram of steel loss. The experimental results confirmed that early age corrosion, verified by internal visual inspection and mass loss measurements, can be detected by AE before any external signs of cracking. Furthermore results show that compressive strength was the primary influential parameter, indicating an exponential, empirical relationship between compressive strength and AE Energy. An increase in temperature usually induces an increase in corrosion activity, which should be measurable using the AE technique. Consequently the influences of seasonal and diurnal temperature variations were investigated to determine their impact on undertaking AE measurements. This phase of the research demonstrated that seasonal variations in temperature impart a negligible influence on measured AE Energy. However measurement of AE Energy per hour followed trends in the diurnal temperature and corrosion rate evolution, these being in a state of constant flux. Therefore AE measurements of corrosion in reinforced concrete are more responsive to a change in temperature, and so corrosion rate, as opposed to a specific and constant corrosion rate. In the final phase practical experience with AE from site trials and laboratory work were coupled with leading research and existing knowledge of corrosion in concrete and structural deterioration, to develop a testing and evaluation procedure for on-site application. This rigorous procedure enables reliable corrosion measurements to be undertaken on reinforced concrete structures using AE technology and enabling an assessment of the rate of corrosion induced damage to be made. As far as the author is aware this is the first site testing procedure for detecting corrosion in reinforced concrete using AE. Future research in this area might involve more site testing with a view to improving accuracy and analysis of on-site data, underpinning the developed procedure.

A study into the performance of CSM (Chopped Strand Mat) reinforced polyester laminates subject to low velocity impact at oblique angles is presented. The investigation encompasses both an experimental and theoretical assessment of the impact event, viz damage tolerance and residual material properties of flat, clamped composite plates. A short introduction is followed by a comprehensive review of the most relevant published literature on all aspects of impact induced damage. A theoretical analysis based on the use of damage toughness parameters is formulated to predict the residual tensile strength and stiffness properties of impact damaged plates. The analysis is further extended to determine the total internal stress distribution in the system. A combination of Hertzian contact, plate bending and finite element solution are used to establish the influence of plate inclination on the resulting stress state. Several common failure criteria were used to predict the extent of the damage. These predictions are compared with experimental data. Results from an extensive experimental programme are presented A fully instrumented test rig was used to assess the influence of imparted energy, impact force, and plate indentation with respect to surface indentation profiles, damage areas and micrographic evidence. The theoretical and experimental results are graphically presented, discussed in detail and exhibit good agreement. The study is finally concluded with comments summarising the most pertinent points derived from the present investigation together with recommendations for further work.

This thesis is concerned with the development of computational procedures in the assessment of the structural integrity and lifetime of cracked bodies subjected to cyclic variable loads and temperatures. The foundation of these techniques is the Linear Matching Method (LMM), related to the methods of elastic compensation and Gloss r-node, used in design calculations for a number of years. It involves matching the behaviours of a non-linear material to that of a linear material, whereby sequences of linear solutions with spatially varying linear moduli are produced. The developed iterative programming algorithms, implemented within the finite element scheme, ABAQUS, would then generate a monotonically reducing sequence of upper bounds, ultimately converging to the least upper bound loads. In their applications, the significance of these programming methods is two-fold. The first is the investigations into the overall behaviour of cracked structures under the combined actions of mechanical and thermal loads. The numerical limit loads and ratchet limits so identified, which describe the onset of plastic collapse and the unlimited accumulation of plastic strains respectively, were found to be stable, with good converged solutions achieved within 40-60 iterations. The analyses also revealed the insensitivity of the ratchet boundaries to cyclic hardening, as the perfectly plastic and complete cyclic hardening limits yielded almost identical results. The other is the examination into the relationship between the near crack tip fields and the cyclic loading histories, in creep and plasticity conditions. It was established that the HRR field criterion is an appropriate representation of the behaviour of the mechanically and thermally induced crack tip fields. This enabled the crack tip fracture criterion to be evaluated in all conditions, with the observed phenomenon described by two distinct behaviours; strongly influenced by the effect of the elastic stress intensity factor and the reference stress respectively. The analyses conducted demonstrated the capability of the adopted numerical procedures in appraising the behaviour of cracked structures under cyclic loading histories, with the conservativeness of current solution procedures in R5 clearly evident in the results enclosed.

The work presented in this thesis is concerned with the static and dynamic analyses of single lap-jointed cantilevered beams. An analytical model is presented for predicting the stress distribution within a single lap-jointed beam under tension. With suitable boundary conditions, the stress distribution of the adhesive at the longitudinal direction is determined. However, the simplifying assumptions of the analytical model are too restrictive. Consequently, all analysis carried out are based on the finite element method. The stress distribution of a single-lap jointed cantilevered beam under tension has been investigated using the three dimensional FEA method. The results show that there is considerable variation in the stress distribution across the width, and that the highest stress concentrations occur in the portions of the adhesive layer closest to the clamped end of the bonded beam. Four finite element models to analyse the behaviour of adhesively bonded joints are described. Comparisons are performed between models with different modelling approaches as well as different types of element combinations in order to find a suitable model to predict the mechanical behaviour of adhesively bonded joints. The effects of typical boundary conditions and of adhesive properties were also investigated. The results indicate that the stress distributions of a single-lap jointed cantilevered beam are strongly affected by the boundary conditions of the beam and by the elastic properties of the adhesive.

The problem of Dynamic Structure-Soil-Structure Interaction (SSSI) refers to the mutual interaction of adjacent buildings in built-up high density areas through the underlying soil under earthquake excitation. Due to the complexity of the problem, past studies have mainly considered the use of intricate mathematical formulations or the computationally demanding numerical Finite Element and Boundary Element methods. In the present study, linear elastic two-dimensional formulations are proposed using simple discrete lumped parameter models for structures and soil for groups of two and three adjacent buildings systems. The formulation includes a rotational spring as a key buildings interaction mechanism. Inverse power laws are proposed for this rotational interaction and for soil/foundation springs stiffnesses which turn out to be functions of spacing between adjacent buildings. These relationships are obtained by equating energies from the low order discrete and high order Finite Element models.

Buildings with non coincident centres of mass and stiffness respond in both translation and rotation during seismic ground excitations. This translational and rotational interaction (torsional coupling) can lead to excessive forces in some structural members. This could possibly lead to structural failure if the building is not properly designed to accommodate this response interaction. Previous elastic analytical studies have determined the structural parameters that govern the degree of torsional coupling. However, the parameters found influencing torsional coupling during inelastic response in previous analytical studies have been found to be both more numerous and contradictory than those associated with the elastic response. This study concentrates on the inelastic behaviour of a series of four storey models representing idealized buildings. These building models have been developed from a previous experimental study on the elastic behaviour of torsional coupling. In this inelastic study, hinge units have been designed and used to simulate the yielding of the column or beam members in the experimental model, while maintaining ease of repeatability between tests. The yielding moment in these hinge units can be adjusted to alter the effective strength of the columns or beams in the model. This, along with the ability to vary the floor mass distribution, column sizes (diameter and length), and stiffness distribution allows for a degree of control on the structural parameters deemed important in previous inelastic analytical studies. Results are presented which illustrate the effects that the various structural configurations have on the different measures of inelastic building response, and its vulnerability to damage. These include changes in the building frequencies, member displacement ductilities and vulnerability, hysteretic energy dissipation, and peak structural responses. The study presents a comprehensive investigation of the column-yielding building models. Additionally, select key cases of the column yielding configurations are compared to both the beam-yielding models, and a computational model.

The main objective of this study was to bring together two areas of research the first being a laboratory study of the time-dependant properties of concrete while the second is a study of the time-dependant behaviour of two prestressed concrete structures - the Grangetown and Cogan Viaducts. A study of the influence of shrinkage and creep on a range of normal to high strength concretes in the laboratory found that as concrete strength increased, drying and total shrinkage also increased while the rate at which creep developed decreased. It was shown that autogenous shrinkage is more prevalent in high strength concrete, and the inclusion of pfa in concrete reduces the amount of total shrinkage. The benefits and limitations of using current shrinkage and creep prediction models were determined through a sensitivity study of their input parameters and through comparisons with shrinkage and creep strains determined in the laboratory. It was found that certain models are more sensitive to specific parameters than others. All models predicted strains which gave good to adequate agreement with the measured strains when the material parameters were within the ranges specified by the models, but the prediction was less reliable when the parameters used in the models exceeded the recommended values. It was shown that the prediction of the time-dependant behaviour of the Grangetown and Cogan Viaducts using these models is inadequate due to influences not considered by the models such as prestressing forces and the construction sequence. The influence of the construction sequence and prestressing data was investigated using a complex computer code in conjunction with these models, and while it was not possible to improve the efficiency of predicting reliable strain behaviour over time, an appreciation of the level of detail necessary to do this was gained. Finally, the influence of environmental conditions on the strain behaviour of these two structures was shown to be greater than anticipated.

Durability requirements have become a major issue in the design of concrete structures today. The hardening process plays a key role in the quality of the concrete. Autogenous shrinkage is considered to be a factor that may cause damage to the concrete structure during hardening. The concept of autogenous shrinkage is relatively new and in the case of conventional concrete with fairly high water to cement ratios these self-induced volume changes are found to be relatively small and therefore are neglected. Self-desiccation and autogenous shrinkage are pronounced phenomena in the case of low water to cement ratio concretes. The current study deals with the development of autogenous shrinkage strains in a new class of High Performance Fibre Reinforced Cementitious Composites (HPFRCCs) designated CARDIFRC that has been recently developed at Cardiff University. The scope of the study was to investigate how the self-induced shrinkage strains develop in CARDIFRC matrix without fibres and what was the effect of the inclusion of a large amount of fibre on the autogenous shrinkage. Both experimental and theoretical studies were undertaken as a part of this investigation. Autogenous shrinkage strains were measured on large and small prisms of CARDIFRC under isothermal conditions. The experiments revealed a relatively large scatter in the measured values for the case of large beams with fibres, whereas the beams of same size but without any fibres gave consistent results. This large scatter has been confirmed to be a result of the uneven distribution of fibres in the large prisms. Small prisms with and without fibres gave very consistent results, with autogenous shrinkage taking place up to 75 days. The autogenous shrinkage strains have been modelled using a thermodynamic approach which follows the continuous change in the moisture content, pore volume and stiffness of the mix with degree of hydration. The predictions of the model are in good agreement with the measured strains in all specimens with and without fibres.

The flexural fracture and fatigue response of high performance fibre reinforced cementitious composites, designated CARDIFRC , was investigated in this study. CARDIFRC is characterised by high tensile/flexural strength and high energy absorption capacity (i.e. ductility). The special characteristics of CARDIFRC make it particularly suitable for repair, remedial and upgrading activities (i.e. retrofitting) of existing concrete structures. One of the major factors affecting the flexural fracture and fatigue behaviour of CARDIFRC specimens was found to be the distribution of fibres within the mix, and it is on this factor that the greater part of this thesis is focused. An even and proper distribution of fibres can lead to excellent flexural fracture behaviour and an extremely high fatigue life of CARDIFRC specimens. On the other hand, poor fibre distribution results in undesirable performance and failure, well below the designed capacity of the material. In particular, the thesis addresses the following key points. The first point concerns the static flexural behaviour of CARDIFRC specimens and the prediction of their load-displacement behaviour. A nonlinear cracked hinge model has been used for the simulation, and the analytical results were found to be in very good agreement with the test results. In addition, a combined damage/fracture mechanics approach to the description of the flexural behaviour is presented, in which a continuum damage model is used up to the peak load followed by a fracture mechanics approach when the damage has localised along the eventual fracture plane. The second key point concerns the flexural fatigue behaviour of CARDIFRC , subjected to several stress amplitude ranges. Careful preparation of CARDIFRC specimens guarantees an excellent and absolutely consistent fatigue response. The endurance limit of the material is very high, not very often observed in the relevant literature. This is an indication that CARDIFRC has an excellent flaw tolerance. Despite the excellent fatigue performance, some small internal damage was noticed, after a very large number of cycles, but this damage is distributed, without the development of any visible crack on the surface. The last key point concerns the distribution of a large volume of short steel fibres in the mix. A thorough investigation of the fibre distribution was accomplished by image analysis of selected planes of failure, in order to explain the corresponding specimen fatigue behaviour. Moreover, a near linear correlation between the image analysis data and a novel non-destructive technique based on computerised tomography (CT) imaging was observed.

A new retrofitting technique using CARDIFRCRTM, a material compatible with concrete, has recently been developed at Cardiff University. It overcomes some of the problems associated with the current techniques based on externally bonded steel plates and FRP (fibre-reinforced polymer), which are due to the mismatch of their tensile strength and stiffness with that of concrete structure being retrofitted. This study investigates the effect of thermal cycles on the performance of reinforced concrete control and retrofitted beams. The concrete beams were heated to a maximum temperature of 90°C from the room temperature of about 25°C. The number of thermal cycles varied from 0 to 90 cycles. After the requisite number of thermal cycles, the beams were tested at room temperature in four-point bending. The tests indicate that the retrofitted beams are stronger, stiffer and more importantly failed in flexure. No visual deterioration or bond degradation was observed after thermal cycling of the retrofitted beams (the bond between the repair material and the concrete substrate remained intact) attesting to the good thermal compatibility between the concrete and CARDIFRCRTM. Therefore, this type of retrofit material can be successfully used in hot climates. The study also evaluates the performance of normal and high strength concretes repaired with CARDIFRCRTM using the wedge splitting test (WST). The main factors that could affect the bond between the repair material and concrete such as the surface roughness and thermal cycling are also investigated. It is shown that surface roughness plays a significant role in the overall bonding system, and no visual deterioration is observed after thermal cycling. Two analytical/computational models for predicting the ultimate moment capacity and the complete load-deflection behaviour of the retrofitted beams were applied. Both models predict very well the ultimate moment capacity of the retrofitted beams.