The research described in this dissertation follows on from the study made by Jewell (1980)into the effects of tensile reinforcement on the mechanical behaviour of sand. For this study Jewell used the direct shear test with reinforcement placed about the central plane as shown in fig. 1.1. The direct shear test was chosen for the following reasons. (1) The reinforcement variables could be better controlled and examined in a unit cell test than in modular field studies of soil reinforcement systems. (2) The pattern of deformation is similar to that experienced by soil in which a rupture band develops, with the principal axes of stress, strain and strain increment free to rotate as is the case in model and field structures. (3) The overall shear strength of the sample is measured directly at the boundaries of the apparatus. The direct shear tests were monitored by boundary measurements and internal measurements using a radiographic technique. The findings are outlined below with reference made to relevant observations by other researchers. 1) The optimum orientation for a relatively flexible steel grid was found to be approximately along the direction of principal tensile strains in the unreinforced sand, see fig.1.2. This indicated that the reinforcement functioned by limiting tensile strains in the sand. McGown et al. (1978) obtained a similar result for plane strain cell tests on sand containing a single layer of flexible reinforcement. However in both studies the reinforcement was observed to waken the sand. Jewell recognized weakening to occur when the steel grid was placed along the direction of principal compressive strains in the unreinforced sand. This was attributed to a reduction in vertical effective stress. McGown et al. observed weakening of the sand when the reinforcement orientation approached the rupture band which developed in the sand alone. This was recognized to be the direction of zero-extension in the unreinforced sand. The weakening was linked to a lower bond between soil and reinforcement than soil alone. 2) Internal strains determined by Jewell showed the tensile reinforcement modified strains in the sand over a well defined zone, see fig.1.3. This resulted in a significant rotation of principal axes of strain increment, with the bond of major strains which developed across the centre of the box in the unreinforced sand being prohibited from forming. This agreed with boundary measurements, indicating the reinforcement functioned by limiting tensile strains in the sand. Consequently a less favourable mode of failure took place. The limit of rotation of principal axes of strain increment was understood to be the alignment of a direction of zero-extension in the sand with the reinforcement. These findings agree with the ideas expressed by Basset and Last (1978) on the mode of action of tensile reinforcement, which in particular was related to the effect of tensile reinforcement on the strain field in a reinforced earth wall as shown in fig.1.4. 3) For efficient use of tensile reinforcement it was demonstrated that the bond with sand should be as high as possible. This could be achieved by roughening the surface. Alternatively, the bond was improved by introducing openings or apertures in the reinforcement, changing the shape to a grid. It appeared that the bond for a suitably proportioned grid could be as high as for a fully roughened surface. 4) The longitudinal stiffness of tensile reinforcement was observed to affect the magnitude and rate of increase in strength in the direct shear tests. The rupture strain of tensile reinforcement relative to maximum tensile strains of the soil, under the same operational stress conditions, have also been observed to influence the reinforcing effect in terms of its limiting behaviour, i.e. whether brittle or ductile (McGown, et al. 1978). With regards to the performance of reinforced earth walls, Al-Hussanini and Perry (1976) observed that steel reinforced strips produced a stiffer and stronger structure than a more extensible fabric reinforcement, even though surface roughness was less. The importance of reinforcement tensile stiffness is recognized in limit equilibrium designs for tensile reinforced soil structures by limiting the available reinforcement force to the tensile strains that can develop in the soil (e.g. Jewell 1985). For highly structured non-woven and composite geotextiles, McGown et al. (1982) demonstrated that the stress-strain behaviour can be significantly affected by soil confinement. Testing wider strips in isolation was not found to replicate the effects of soil confinement. Another factor which needs to be considered when assessing the tensile property of a polymer reinforcement is creep. McGown et al. (1984) illustrated an appropriate method of interpreting creep data using isochronous curves, which enable long term laboratory test data to be extrapolated to the design life of the soil structure. 5) The strain and hence stress fields in the reinforced direct shear tests have been shown to be complex and non-uniform. However Jewell successfully modelled the variation of reinforcing effect for tensile reinforcement at different orientations by using a simple limit equilibrium analysis, see fig.1.5. The effect of the tensile reinforcement force was represented as: - an increase in the normal effective stress acting on the central plane of the box due to the normal component of the force and - a reduction in the applied shear stress due to the parallel component of the force to the central plane. Subsequently this analysis has been applied to limit equilibrium design methods for reinforcing soil retaining walls and embankments, Jewell et al. 1984, and Jewell 1982 respectively. 6) A reduction in the reinforcing effect for individual reinforcement due to the presence of other reinforcement was observed in the shear box. This loss of efficiency of individual reinforcement was termed interference. Interference between tensile reinforcement has also been studied by Guilloux et al. (1979) for the pull-out resistance from soil. However interference between reinforcement has yet to be introduced into a limit equilibrium design method.

Pipe jacking is a trenchless construction technique for the installation of underground pipelines. Although pipe jacking is widely used, fundamental research is still needed to understand fully the factors affecting the process and to prevent unexpected failure. With the time and financial limitation, it is difficult to explore all aspects of these factors with experiments; and it is also difficult to study them by analytical methods because of the complexity of the problem. This thesis describes the use of the finite element technique to study the pipe performance under different environments and the laboratory tests of several different joint designs. The emphasis of the current research is on the performance of the concrete pipes during jacking under working conditions and to seek possible improvements in the design of pipes and pipe joints by numerical modelling. In the finite element modelling, a simplified two-dimensional model is used for a preliminary study, then the analyses are carried out with three-dimensional models A, B and C representing a complete pipe, a pipe with surrounding soil and a symmetric three-pipe system respectively. Several factors affecting the pipe performance have been examined, for example, the properties of the packing material, the stiffness of the surrounding soil, the misalignment angle at the pipe joint, and the interaction between the pipe and surrounding soil. The numerical results show that the misalignment of the pipe is the dominating factor inducing both tensile stresses and localized compressive stresses in the concrete pipe, especially with a high misalignment angle which results in separation between packing material and the pipe. The packing materials with high Poisson's ratio and high stiffness also induce higher tensile stresses in the pipe, and the influence of Poisson's ratio is significant. Under 'diagonal' loading, both the stiffness of the surrounding soil and the interaction between the pipe and the surrounding soil have a significant effect on the stresses within the concrete pipe. Under 'edge' loading, the greatest potential damage is at the pipe joint due to the tensile stresses in the hoop direction; while under 'diagonal' loading, the damage is most likely the cracking on the external surface of the pipe along a line connecting the two diagonal loaded corners. The results show that the Australian model gives somewhat good prediction about the maximum normal stress and the diametrical contact width at pipe joint. Based on the numerical results, several different joint designs for improving the pipe strength have been proposed and tested in the laboratory. Both the laboratory tests and the back analyses suggest that the local reinforcement and the local prestressed band at the pipe joint will improve the pipe strength.

The mechanisms by which a geogrid acts to reinforce a granular layer over a soft clay were investigated with reference to trafficking of unpaved roads. The research involved a programme of physical testing at both model and full scale. Model tests were carried out for monotonic and cyclic loading of a dual footing on to a layer of granular material that was compacted over a consolidated kaolin sample. The tests were made in conditions of plane strain at a quarter scale, and included a scaled geogrid at the base of the granular layer. Measurements of footing load, displacements and photographic observations through a perspex front face to the test box were used to identify the reinforcing actions of the geogrid. A virtual work procedure was adopted in analysis of the results for both unreinforced and reinforced layers. The procedure accounts for lateral restraint of the subgrade and a membrane action in the reinforced tests, and allows a comparison of the cyclic test results with an equivalent monotonic bearing capacity. Load tests were also made on footing plates at full scale, involving a similar arrangement of compacted granular layers over prepared London clay and including a high strength polymer geogrid. Examination of the test results allowed a qualitative assessment of the model performance and the reinforcing mechanisms attributed to the geogrid.

An instrumented model pile was used to investigate the fundamental behaviour in clay soils of driven cylindrical steel piles used for offshore structures. Four test-bed sites were chosen; two in stiff heavily consolidated clays, and two in normally/lightly overconsolidated clays. Data from these sites confirm that a residual shear surface is formed along the pile during installation, the location of which relative to the shaft surface appears to depend on the shaft roughness. Comparisons with other site investigation data and cavity expansion theoretical predictions indicate that stress relief immediately behind the pile tip during driving gives rise to total radial stresses and pore pressures measured on the pile shaft which are lower than predicted. This stress relief is particulary severe in the stiffer clays. The data did however show that the installation total radial stresses and pore pressures are governed by the initial in-situ stresses and undrained shear strength as is predicated by the theory. During reconsolidation, pore pressures close to the instrument rise initially in all clays, and radial effective stresses drop. The slow recovery in radial effective stress during the later stages of reconsolidation was in some cases insufficient to return it to levels recorded during installation. However, the generation of negative pore pressures during undrained loading increased the radial effective stress and shaft friction at failure. This effect is particularly important in the normally consolidated clays, and is responsible for the set-up of shaft capacity seen in such clays, which might not be observed if the loading were drained. The observed behaviour was therefore quite different from the monotonic increase in radial effective stress during reconsolidation, followed by decrease during undrained loading which was expected from a review of current theory.

The development and implementation of a complex numerical model for the determination of the damage to masonry buildings resulting from tunnelling settlements is described in this thesis. The current methods of damage prediction do not, in general, take into account the stiffness and weight of the surface structure. The model addresses this deficiency by explicit inclusion of the structure. Three-dimensional finite elements are used to model the ground with a non-linear, elasto-plastic soil model based on kinematic hardening. Tunnel linings are modelled using a novel overlapping elastic shell element: volume loss being simulated by shrinkage of linings coincidentally with excavation. Structures are modelled as collections of facades comprised of plane stress elements using a non-linear material model for masonry, similar to elastic-no tension. In developing the three-dimensional model, its two-dimensional counterpart is also studied. While the beam and shell elements used for linings (in two- and three-dimensions respectively) have the advantage of no rotational degrees of freedom the need to model boundary conditions at the element stiffness level complicates implementation. Tests using the shell elements show them to be satisfactory for the purpose of modelling tunnel linings. Results from a small number of analyses are given for construction of a straight tunnel beneath simple masonry structures. It is shown that the effect of the building on settlements depends heavily on its location in plan with respect to the tunnel axis. Predictions of crack patterns using the model for these analyses show that facades which the tunnel passes under first are less damaged than those later in the excavation sequence. Both of these conclusions serve to demonstrate that the problem can only be realistically modelled using three-dimensional methods. At present, however, the computer resources required to run the three-dimensional model are considerable.

The reported work constituted the third phase of a programme of research into the performance of concrete pipes during installation by the pipe-jacked tunnelling technique. This third stage was a continuation of the on-site monitoring of full-scale pipe jacks during construction. Four schemes were monitored in different ground conditions: London clay, dense fine sand below the water table, stiff glacial till and soft alluvial clay. Pipe sizes ranged from 1000mm to 1800mm internal diameter and excavation methods included hand tools, slurry machines and an open face tunnel boring machine. The main objective was to collect information on jacking loads and stresses at the pipe-soil interface to provide a better basis for future designs. This was achieved by building twelve stress cells -capable of measuring total normal stresses, shear stresses and pore pressures - into the wall of a standard concrete jacking pipe that could be inserted anywhere in the pipe string. Jacking loads and forward movement of the pipe string were simultaneously recorded and the results were correlated against site activities, including lubrication operations, and tunnel alignment surveys. Another objective was to monitor the ground response by measuring displacements around the tunnel and ground pressures above the perpendicular to the intended line. Ground movements were measured using conventional surveying techniques for surface settlements, and inclinometer access tubes for sub-surface deformation. On one scheme, electro-levels were employed in a near-horizontal tube to measure centre line settlement as the tunnel bore advanced. Push-in spade cells and pneumatic piezometers were installed on two schemes to measure the change in horizontal pressures with the passage of the shield. Because of the myriad data collected, it has only been possible to present a summary of the results obtained. Jacking force records from all the monitored schemes - including the previous fieldwork stage - are presented. The pattern of jacking load build up and the magnitude of frictional resistance can differ significantly according to the type of ground and use of lubricants. Stress measurements at the pipe-soil interface show that the interaction between jacking loads, pipeline misalignment, stoppages, lubrication, excavation technique etc, is highly complex. Ground movement measurements compared to well established empirical predictive methods show that short-term displacements are related to ground losses caused by closure of the overbreak void between shield and pipe.

Currently, much research is being directed at alternative energy sources to supply power for modern life of today and the future. One of the most promising sources is wind energy which can provide electrical power using wind turbines. The increase in the use of this type of energy requires greater consideration of design, installation and especially the cost of offshore wind turbines. This thesis will discuss the modelling of a novel type of shallow foundation for wind turbines under combined loads. The footing considered in this research is a circular caisson, which can be installed by the suction technique. The combined loads applied to this footing will be in three-dimensional space, with six degrees of freedom of external forces due to environmental conditions. At the same time, during the process of building up the model for a caisson, the theoretical analyses for shallow circular flat footing and spudcans also are established with the same principle. The responses of the soil will be considered in both elastic and plastic stages of behaviour, by using the framework of continuous plasticity based on thermodynamic principles. During this investigation, it is necessary to compare the numerical results with available experimental data to estimate suitable values of factors required to model each type of soil. There are five main goals of development of the model. Firstly, a new expression for plasticity theory which includes an experimentally determined single yield function is used to model the effects of combined cyclic loading of a circular footing on the behaviour of both sand and clay. This formulation based on thermodynamics allows the derivation of plastic solutions which automatically obey the laws of thermodynamics without any further assumptions. A result of this advantage is that non-associate plasticity, which is known to be a proper approximation for geotechnical material behaviour, is obtained logically and naturally. A FORTRAN source code called ISIS has been written as a tool for numerical analysis. Secondly, since there are some characteristics of the geometric shape and installation method which are quite different from that of spudcans and circular flat footing, another objective of this study is to adapt the current model which has been developed in ISIS for spudcans to the specific needs of caissons. The third goal of this research is the simulation of continuous loading history and a smooth transition in the stress-strain relationship from elastic to plastic behaviour. The model is developed from a single-yield-surface model to a continuous plasticity model (with an infinite number of yield surfaces) and then is discretized to a multiple-yield-surface model which can be implemented by numerical calculation to be able to capture with reasonable precision the hysteretic response of a foundation under cyclic loading. This can not be described by a conventional single-yield-surface model. Fourthly, as a method to simplify the numerical difficulties arising from the calculation process, a rate-dependent solution will be introduced. This modification is implemented by changing the dissipation function derived from the second law of thermodynamics. Finally, in order to control the model to capture the real behaviour, many parameters are proposed. A parametric study will be implemented to show the effects of these parameters on the solution.

The construction of underground tunnels in soft ground in urban areas involves the potential for ground movements caused by the tunnelling to affect existing surface structures. Masonry structures are at particular risk of crack damage. Conventional empirical building assessments do not fully capture all aspects of this soil-structure interaction situation. Numerical methods are increasingly used for such problems. It is common practice in empirical and numerical methods to model a building as an elastic beam in 2D. The objective of this thesis is the development of a new approach to the numerical modelling of masonry buildings using surface beams in 3D. In phase one of this project, finite element analyses of elastic and masonry facades are undertaken and the traditional beam method of modelling them is assessed. New equivalent elastic surface beams are developed, the properties of which account for the dimensions and openings in facades which were found to influence the response to settlements. Equivalent masonry beams are also developed which have a constitutive model that accounts for the different response of masonry buildings in hogging and sagging. Timoshenko beams are chosen to model the facades and these beams were implemented into the OXFEM finite element program with full 3D capability along with the new constitutive beam models. Example masonry structures were modelled in 3D using the new surface beams in phase two. Tunnel construction was simulated under the buildings and the response of the beams compared to a full masonry building model. Example analyses included buildings both symmetric and oblique to the tunnel. Results showed that the equivalent elastic beams accurately simulate full masonry building response in sagging regions. Parametric studies confirmed the choice of equivalent beam parameters and the impact of different relative stiffnesses. The equivalent masonry beams displayed the same good agreement in sagging but were less accurate in hogging. In phase three, finite element models are used to compare ground movements and structural response of buildings using the 3D equivalent masonry beam method and observed data from the construction of the London Underground Jubilee Line Extension. The surface beams showed good agreement with the observed building responses in both sagging, where the building response was essentially rigid and in hogging where a more flexible response was observed.

The stress-strain behaviour of unsaturated soils containing discrete bubbles of gas has been studied in a programme of experimental and theoretical research. The research is of particular relevance to the offshore environment, where bubbles of methane, nitrogen and carbon dioxide are formed within the seabed by the decomposition of organic matter. The presence of these gas bubbles can have a major influence on the engineering properties of the soil. As gas bubbles in marine sediments are typically much larger than the normal void spaces, the bubbles cannot be considered as simply changing the compressibility of the pore fluid. Chapters 2 to 4 describe a series of undrained triaxial tests on reconstituted samples of clayey silt from an estuarine site. Bubbles of methane were formed within the soil by using a chemical technique. The test results provide evidence of the effect of gas bubbles on the undrained shear strength, together with useful information on the stress-strain behaviour prior to failure, the generation of pore pressures during shearing, the elastic moduli and the isotropic consolidation behaviour. A theoretical model for soils containing large gas bubbles is developed in Chapter 5. The model consists of spherical bubble cavities surrounded by a matrix of saturated soil. In the following three chapters various aspects of the soil behaviour are examined by analysing the theoretical model. Chapter 6 covers the elastic moduli, Chapter 7 the consolidation behaviour and Chapter 8 the undrained shear strength. In Chapter 9 the predictions of the theoretical model are compared with the experimental results. The agreement between theory and experiment is excellent, suggesting that the theoretical approach developed and analysed in Chapters 5 to 8 is a reasonable model for the behaviour of soils containing large gas bubbles.

This thesis presents the results of an investigation into soil-reinforcement interaction by means of direct shear and pull-out tests. Scale and other factors affecting test results were studied; for this purpose an apparatus able to contain a 1 cu.m sample of sand was designed by the author in order to perform large scale tests. Plastic and metal sheet and grid reinforcements were used in conjunction with Leighton Buzzard sand. Direct Shear tests on unreinforced sand samples showed that soil strength parameters were not affected by the test scale, although the post peak behaviour and the shear band thickness at the centre of the sample were significantly affected by the scale of the test. The presence of a reinforcement layer inclined to the central plane of the box had a marked effect on the strength and behaviour of the sample. The reinforcement increased the vertical stress and inhibited the shear strain development in the central region of the sample. The behaviour of the reinforced sample was found to depend on the type and form of the reinforcement as well as its mechanical properties. Pull-out test results can be severely affected by boundary conditions, in particular by the friction on the front wall of the box. The results obtained in the series of tests showed that interference between grid bearing members is the main factor conditioning the pull-out resistance of a grid reinforcement. The intensity of such interference was quantified on the basis of results obtained in tests using single isolated bearing members and grids with different geometric characteristics. An expression for the bond coefficient between soil and grid, taking into account the degree of interference, was suggested. It was also observed that the maximum bearing pressure exhibited by a bearing member depends on the ratio of the member diameter to the mean particle size.