A new set of laboratory experiments to examine the short-term statistics of crest elevation and wave heights has been undertaken. Sea states with a range of steepness and directional spreading have been considered. Comparisons between these data and a number of widely adopted short-term statistical models exhibit clearly defined departures. For a given sea state, the extent of these departures is directly proportional to the sea state steepness and inversely proportional to the directional spread. With directional spreading identified as a critical parameter, a detailed study of how best to describe, define and model it has been undertaken. The key finding of this study is that the average directional spread in the steepest sea states reduces. In addition, it has also been shown that on average the largest waves in these steep sea states are more uni-directional when compared to the sea state as a whole. Further consideration of the data show that the two physical mechanisms leading to the alteration of the statistics are nonlinear amplification (leading to increases above second-order) and the dissipative effect of wave breaking. Quantification of the effects arising from these two competing mechanisms has been undertaken based on additional simulations (both numerical and experimental) of focused wave groups. For uni-directional sea states, a classical expansion (truncated at a third order of wave steepness) of the increased surface elevation obtained in a fully nonlinear uni-directional focused wave group has been used to quantify the effect of amplification in the crest height statistics. Similarly, the dissipative effect of wave breaking on crest elevations has been quantified based on the reduction in crest elevations in focused wave groups with linear amplitude sum larger than the limit at which incipient spilling first occurs. These reductions are calculated as the difference in the maximum crest elevation in a breaking wave event and that predicted by the third-order power series used for the quantification of nonlinear amplificiation. Overall the two methods employed in quantifying the effect of nonlinear amplification and wave breaking yield good agreement with the original (random) laboratory data. Finally, directionality is incorporated into these predictions based on the linear reduction in the wave front steepness with increasing directional spread. Both the nonlinear amplification and the dissipative effect of wave breaking are calculated based on this reduced steepness for the directional sea states. The predicted crest heights from this simplified procedure compare well with the laboratory data; the predictions remaining conservative throughout.

Concentrically braced frames are a common form of earthquake resistant structure. Performance of the structure is largely dependent on the ability of the key dissipative components, in this case the diagonal bracing members, to undergo significant inelastic deformations. Whilst many earlier studies have examined the hysteretic response of bracing members, comparatively less attention has been given to the ultimate behaviour and failure conditions. There are also significant uncertainties in existing models for predicting the ductility capacity of braces owing to their semi-empirical nature as well as the scatter of test results. This research project examines the cyclic behaviour of tubular braces made of a familiar structural material, carbon steel, and an increasing popular alternative structural material, stainless steel, which is known for its high tensile ductility. As part of the current study, laboratory tests were performed on hot-rolled carbon steel, cold-formed carbon steel and cold-formed austenitic stainless steel hollow section members and materials coupon cut from them. A total of 24 tensile coupon tests, 62 cyclic material tests and 16 cyclic member tests were conducted. Strain-life relationships of the materials under low and extremely low cycle fatigue regimes were established from the results of the cyclic material tests. These data were also used for calibrating material cyclic hardening models, which were incorporated in numerical models of hollow section members. These models, verified against the results of the cyclic member tests from this study and other research programmes, were employed in parametric studies to investigate the influence of member geometry and material properties on the behaviour of the bracing members. Although the three materials exhibit similar strain-life relationships, cold-formed stainless steel members perform better in terms of displacement ductility and energy dissipation, which is due to the cyclic hardening and higher post-yield stiffness of the stainless steel material. Implications of these findings on the design of earthquake resistant concentrically braced frames are discussed and design guidance for stainless steel bracing members is proposed.

The microstructures of Jurassic and Cretaceous mudrocks reflect their environment of deposition and the processes which occurred during their burial. One measurement of mudrock microstructure is particle long axis orientation. Preferred particle orientations are measured by analysis of SEM images of broken mudrock surfaces. Strong preferred particle orientations are observed in unweathered material with high quantities of illite. They are observed across a range of burial depths within Cretaceous and Jurassic materials. No preferred orientation is observed where a material is weathered, or where large quantities of silt are present. Moderate preferred orientations are observed where materials are carbonate rich such as with the Blue Lias, or are shallowly buried such as with the London Clay. Strong preferred particle orientations formed through particle rearrangement are created at greater burial depths than the London Clay samples (200m) and at shallower depths than the Gault Clay samples (500m). Microstiuctures form in response to each event in a mudrocks history. The final structure of a mudrock is dependent on all of the events in its history. The relative effect of each event is dependent on its intensity of its application to the mudrock. Each event has the possibility of destroying structures formed by previous events. For example burial microstructures often replace depositional structures. Particles will align by rearrangement as much as possible during initial consolidation. Further alignment can occur by the process of pressure solution at a much greater depth. Fluids rich in chemical ions may deposit cements in the pores of mudrocks as they flow through them. The growth of these cements can make further alignment of clay mineral particles impossible.

Reinforcement spacers (i.e. bar supports, chairs) are crucial elements of reinforced concrete, but their influence on the microstructure and long-term durability is not clear. This study investigates the effect caused by plastic and cementitious spacers, and steel wire chairs combined with different aggregate sizes, curing and conditioning regimes on the transport properties, microstructure and chloride-induced corrosion of concrete structures. Concrete cylindrical samples were prepared with 25 and 50 mm high plastic, steel and cementitious spacers. Samples were then cured, conditioned and tested for oxygen diffusivity, oxygen permeability, water sorptivity and chloride diffusivity. Selected samples were pressure impregnated with fluorescent epoxy to study the extent and spatial distribution of epoxy intrusion. The interfacial zone between the spacer and concrete was examined using field-emission scanning electron microscope in the backscattered electron (BSE) mode. The ingress of chloride, particularly near the interface between spacers and concrete matrix was studied using micro X-ray fluorescence (μXRF). The effect of plastic and cementitious spacers on chloride- induced corrosion via capillary rise and cyclic wetting/drying was investigated using small reinforced concrete beams. The feasibility of improving the bond between spacer and concrete by increasing surface roughness of plastic spacers was also investigated. Results show that concrete samples containing plastic spacers consistently gave the least resistance to transport and the highest epoxy penetration followed by samples with cementitious spacers, and then steel spacers. The control samples (samples without spacers) had the highest resistance to transport in all cases. The epoxy penetration occurred mainly through the spacer-concrete interface. The microstructure of the spacer-concrete interface showed significantly lower cement content and higher porosity compared to 'bulk paste' farther away form the interface. Higher penetration of chloride ions was detected along spacer-concrete interface compared to the control sample or the bulk paste farther away. It is evident that spacers initiate early corrosion and this may reduce the service-life of reinforced concrete structures. The implications of these findings on durability of concrete structures are discussed. Several recommendations to improve the bond at the interface between spacer and concrete are presented.

This thesis investigates the implications of incorporating active vibration control (AVC) into floor structures from the initial design stage, with the goal of enabling the construction of more slender long-span floors. The original contributions to knowledge in this work are the investigations into: the development of a novel walking force that simulates the in-service loading of an office environment; the comparison between the effectiveness of AVC and tuned mass dampers (TMDs) when used on floor structures; the investigation into the effect of AVC over the entire floor area rather than considering single locations only, leading to conclusions about typical numbers of actuators that would be required; the investigation into the trade-off between power demand and the performance of an AVC system; and the initial life cycle analysis (LCA) of a floor that incorporates AVC at the design stage. The force model utilises simultaneous pedestrians walking throughout the structure and was calibrated and verified using experimentally acquired data. AVC was found to be a significant improvement upon TMDs in that the response of the structure was reduced to a greater extent using a much smaller inertial masses. The effectiveness of AVC was generally limited to within a single bay. However, large reductions in response were observed within each controlled bay. Therefore, it is suggested that a rule of thumb of one actuator per significant panel is required to control a given floor area, and that the size of these bays should be maximised to increase the effectiveness of AVC. High feedback gains resulted in only slight improvements in structural response, therefore improvements in the non-overhead power demand for AVC can be achieved through a simple decrease in the feedback gain. This has the additional benefit that smaller actuators could be utilised. The initial LCA highlighted the high financial cost of AVC but also demonstrated that potentially significant material savings could be realised through incorporation of AVC at the design stage.

After the Northridge 1994 earthquake it was found that failures in steel frame buildings were mainly concentrated in beam-column connections. This prompted researchers to work on improvements, mainly focussed on increasing the rotational ductility capacity in connections. Most of these improvements however did not eliminate the residual deformations in the connections. To overcome this weakness, researchers introduced post-tensioned steel connections, composed of post-tensioned steel strands and energy dissipating devices. In this research, a single-element model of post-tensioned connection was developed and incorporated in a new computer program for non-linear dynamic frame analysis, which was then used to investigate the effects of the level of post-tensioning forces on seismic behaviour of frame buildings. When used in moment resisting frames, post-tensioned connections reduce residual displacements and prevent development of plastic hinges in the beams. The initial stiffness of post-tensioned frames is also similar to conventional moment resisting frames but their energy dissipation capacity is lower. The performance of the structure is sensitive to the level of post-tensioning forces, and in some cases the use of post-tensioned connections leads to increased displacements. The aim of this research was to investigate semi-active control of the post-tensioning forces as an approach for improving the seismic behaviour of multi-storey steel frame buildings. Three control approaches were proposed to improve the dynamic behaviour of post-tensioned frames: (i) energy dissipation approach which aims to increase the energy dissipation capacity of the frame, (ii) stiffness control approach which aims to change the frequency of the frame by softening or stiffening to avoid excitation by major frequency components of the earthquake and (iii) deformation regulation approach which aims to improve the distribution of deformations along the height of the frame. The three control approaches showed different results. Increasing energy dissipation in the connections is not an efficient approach for reducing the frame response, especially when large displacements occur in the early stages of loading. The stiffness control approach showed good performance, reducing both floor displacements and force demand on the elements. The deformation regulation approach also improved the response, providing more uniform inter-storey drift distribution. In general, the research presented here shows that semi-active control can be used to improve the seismic performance of post-tensioned steel frames.

The work of this thesis is centred around Structural Health Monitoring (SHM) and is divided into three main parts. The thesis starts by exploring different architectures of auto-association. These are evaluated in order to demonstrate the ability of nonlinear auto-association of neural networks with one nonlinear hidden layer as it is of great interest in terms of reduced computational complexity. It is shown that linear PCA lacks performance for novelty detection. The novel key study which is revealed amplifies that single hidden layer auto-associators are not performing in a similar fashion to PCA. The second part of this study concerns formulating pattern recognition algorithms for SHM purposes which could be used in the wind energy sector as SHM regarding this research feld is still in an embryonic level compared to civil and aerospace engineering. The purpose of this part is to investigate the effectiveness and performance of such methods in structural damage detection. Experimental measurements such as high frequency responses functions (FRFs) were extracted from a 9m WT blade throughout a full-scale continuous fatigue test. A preliminary analysis of a model regression of virtual SCADA data from an offshore wind farm is also proposed using Gaussian processes and neural network regression techniques. The third part of this work introduces robust multivariate statistical methods into SHM by inclusively revealing how the influence of environmental and operational variation affects features that are sensitive to damage. The algorithms that are described are the Minimum Covariance Determinant Estimator (MCD) and the Minimum Volume Enclosing Ellipsoid (MVEE). These robust outlier methods are inclusive and in turn there is no need to pre-determine an undamaged condition data set, offering an important advantage over other multivariate methodologies. Two real life experimental applications to the Z24 bridge and to an aircraft wing are analysed. Furthermore, with the usage of the robust measures, the data variable correlation reveals linear or nonlinear connections.

The Great Sumatra earthquake and resulting devastating tsunami in 2004 have highlighted the significance of earthquake risk assessment that can effectively include all main associated hazards, such as tsunami. An integrated quantification of earthquake and tsunami risk is challenging due to mathematical and computational issues as well as limited data, and as such it has not been done before in developing countries. In addition, the conventional Probabilistic Seismic Hazard Assessment (PSHA) generally assumes a Poissonian earthquake model with a stationary rate of hazard. However, hazard rates increase with elapsed time since last large earthquake, or when a seismic gap is present along a fault, and thus, a time dependent (non-Poissonian) PSHA model is generally more suitable. The Earthquake Risk Assessment (ERA) Framework developed at the University of Sheffield is extended in this study to account for tsunami and to include a time dependent hazard assessment. The ERA Framework is based on a stochastic approach that utilizes readily available seismological information. Hence, it is suitable for use in developing countries. The extended framework is used to carry out earthquake and tsunami hazard assessments for Sumatra. This study finds that a maximum PGA of 0.65g on bedrock is expected in the area, a value comparable with that found in other recent studies, but higher than the Indonesian seismic code SNI 03-1726-2002 (the code uses a value of 0.3g). In addition, the time dependent model of this study captured the increased rate of hazard in the middle segment of the Sumatra Subduction Zone, which is consistent with the location of the seismic gap. Currently, the seismicity of this region has increased about 2 times above that of the long term rates. An average tsunami height of 20.7 m is estimated for Padang city at 10% probability of exceedance in 50 years. This study estimates that total earthquake and tsunami risks for buildings in Padang are approximately £54.5 million and £30.8 million per annum, respectively. The annual fatality risk due to tsunamis is much higher than that due to earthquakes, which is approximately 2000 and 8 fatalities, respectively. Total earthquake premiums for the existing building stock of Padang are estimated to be 36.4‰, 16.6‰, 8.1‰ and 3.1‰ for UBM, CBM, RCI and steel structures, respectively. For tsunami hazard, the recommended premium rates are 11.7‰, 9.9‰ and 7‰ for UBM, CBM, RCI/steel structures, respectively. For seismically designed structures, the premium rates decrease about 80% and 25% for earthquake and tsunami hazards, respectively. Earthquake insurance rates applied by insurance companies in Indonesia are comparable with those estimated for seismically designed structures (1.9‰ for RC/steel buildings and 4.7‰ for other buildings). Mitigation strategies to minimize the risks are proposed including the enforcement of seismic design provisions for all buildings in Padang, nationwide obligatory seismic insurance for buildings, tsunami vertical evacuations shelters and tsunami energy dissipation efforts (e.g. offshore barriers, coastal vegetation). At least 17 points of tsunami evacuations refuges are proposed for Padang city, to increase the survival probability of the residents in the area.

RC deep beams are key safety critical structural systems carrying heavy loads over short span, such as transfer girders in tall buildings and bridges. Current design provisions in codes of practice fail to predict accurately and reliably the shear capacity of RC deep beams and in some cases they are unsafe. This work aims to develop a better understanding of the behaviour of RC deep beams and governing parameters, and to improve existing design methods to more accurately predict the shear capacity of such members. An extensive experimental programme examining 24 RC deep beams is carried out. The investigated parameters include concrete strength, shear span to depth ratio, shear reinforcement and member depth. To develop a better insight on the distribution and magnitude of developed stresses in the shear span, finite element analysis is also performed. The microplane model M4 is implemented as a VUMAT code in ABAQUS to represent the behaviour of concrete in a more reliable manner and validated against experimental tests on RC deep beams. This model is utilised in a parametric study to further investigate the effect of concrete strength, shear span to depth ratio and shear reinforcement. The experimental and numerical results show that concrete strength and shear span to depth ratio are the two most important parameters in controlling the behaviour of RC deep beams, and that shear strength is size dependent. The analysis also shows that minimum amount of shear reinforcement can increase the shear capacity of RC deep beams by around 20% but more shear reinforcement does not provide significant additional capacity. A lateral tensile strain based effectiveness factor is proposed to estimate the strength of the inclined strut to be used in strut-and-tie model. Additionally, node factors to estimate the developed strength in different type of nodes are proposed. The proposed model is evaluated against a large experimental database and the results show that it yields more accurate and reliable results than any of the existing models. The model is characterized by the lowest standard deviations of 0.26 for both RC deep beams with and without shear reinforcement and accounts more accurately for all influencing parameters.

The aim of this project is to investigate the stability of cut slopes in stiff clay. The findings are subsequently applied to model stabilisation with piles, used to remediate failure of existing slopes and stabilise potentially unstable slopes created by widening transport corridors. Stiff clay is a strain softening material, meaning that soil strength reduces as the material is strained, for example in the formation of a slip surface. In an excavated slope this can lead to a progressive, brittle slope failure. Simulation of strain softening behaviour is therefore an important aspect to model. The interaction of piles and stiff clay cut slopes is investigated using the Imperial College Geotechnics section's finite element program ICFEP. In designing a suitable layout of the finite element mesh, preliminary analyses found the two existing local strain softening models to be very dependent on the size and arrangement of elements. To mitigate this shortcoming, a nonlocal strain softening model was implemented in ICFEP. This model controls the development of strain by relating the surrounding strains to the calculation of strain at that point, using a weighting function. Three variations of the nonlocal formulation are evaluated in terms of their mesh dependence. A parametric study with simple shear and biaxial compression analyses evaluated the new parameters required by the nonlocal strain softening model. The nonlocal results demonstrated very low mesh dependence and a clear improvement on the local strain softening models. In order to examine the mesh dependence of the new model in a boundary value problem compared to the local strain softening approach, excavated slope analyses without piles were first performed. The slope was modelled in plane strain with coupled consolidation. These analyses also investigated other factors such as the impact of adopting a small strain stiffness material model on the development of the failure mechanism and the impact of the spatial variation of permeability on the time to failure. The final set of analyses constructed vertical stabilisation piles in the excavated slope, represented as either solid elements or one dimensional beam elements. The development of various failure mechanisms for stiff clay cuttings was found to be dependent on pile location, pile diameter and pile length. This project provides an insight into the constitutive model and boundary conditions required to study stabilisation piles in a stiff clay cutting. The nonlocal model performed very well to reduce mesh dependence, confirming the biaxial compression results. However, the use of coupled consolidation was found to cause further mesh dependence of the results.