Suction caissons are a new foundation option for offshore wind turbines. This thesis is focussed on the behaviour of suction caisson foundations in sand and in clay during installation, and under subsequent vertical and combined moment-lateral loadings. The research is based on extensive experimental work carried out using model scaled caissons. The analysis of the results allowed the determination of parameters for hyperplasticity models. Model caissons were vertically loaded in loose and dense sands to study in service states and plastic behaviour. Bearing capacity increased with the length of the caisson skirt. The bearing capacity formulation showed that the angle of friction mobilised was close to the critical state value for loose sands and close to those of peak values due to dilation for dense sands. The vertical load increased, though at a lower rate than during initial penetration, after large plastic displacements occurred. A hardening law formulation including this observed behaviour is suggested. In sand the installation of caissons by suction showed a drastic reduction in the net vertical load required to penetrate the caisson into the ground compared with that required to install caissons by pushing. This occurred due to the hydraulic gradients created by the suction. The theoretical formulations of the yield surface and flow rule were calibrated from the results of moment loading tests under low constant vertical loads. The fact that caissons exhibit moment capacity under tension loads was considered in the yield surface formulation. Results from symmetric and non symmetric cyclic moment loading tests showed that Masing’s rules were obeyed. Fully drained conditions, partially drained and undrained conditions were studied. Caisson rotation velocities scaled in the laboratory to represent those in the field induced undrained response for relevant periods of wave loading, a wide range of seabed permeabilities and prototype caisson dimensions. Under undrained conditions and low constant vertical loads the moment capacity of suction caissons was very small. Under partially drained conditions the moment capacity decreased with the increase of excess pore pressure. In clay, vertical cyclic loading around a mean vertical load of zero showed that in the short term the negative excess pore pressures generated during suction installation reduced vertical displacements. The yield surface and the flow rule were determined from moment swipe and constant vertical load tests. The moment capacity was found to depend on the ratio between the preload Vo and the ultimate bearing capacity Vu. Gapping response was observed during cyclic moment loading tests, but starting at smaller normalised rotations than in the field. The hysteresis loop shape obtained during gapping cannot be reproduced by means of the Masing’s rules.

The consolidation behaviour of gassy soil has been studied in a programme of experimental and theoretical research. This research is of particular importance to the offshore geotechnical industry as the presence of gas in the seabed can have a dramatic effect on the material properties of a marine sediment. Initial numerical modelling based on existing unsaturated soil theory combining the gas and the water phase into a compressible fluid in the pores of compressible soil skeleton failed to simulate the soil behaviour previously observed experimentally at Oxford. Therefore, there was scope for further study in this field. Chapters 2 to 4 describe the experimental preparation, consolidation technique and experimental results of the two series of tests on artificially prepared gassy soil samples. The results of these tests indicated that the gas appeared to be affected by the total stress rather than the pore water pressure, with the saturated soil matrix outside the gas voids being controlled by the consolidation stress. Chapter 5 presents the one-dimensional numerical modelling of the experimental results. Poor simulations were again made using compressible fluid theory. Treating the gas as compressible solid inclusions embedded in a saturated soil matrix, however, resulted in excellent simulations of the observed pore water pressures and settlements. Chapter 6 attempts to explain the results of the experimental and numerical modelling in terms of elastic and plastic soil behaviour. This includes the introduction to the double compressibility model in which the deformation behaviour of the saturated matrix is governed by changes in consolidation stress, whereas that of the gas is governed by changes in total stress. Chapter 7 presents the development of the governing gassy soil consolidation equations under both plane strain and axisymmetric conditions. Chapter 8 describes the approximation of the governing consolidation equations using the Galerkin finite element method in terms of nodal displacements and pore water pressures. The resulting finite element approximation is subsequently formulated for rectangular elements under plane strain and axisymmetric conditions in Chapter 9. The remainder of the thesis describes the structure of the finite element model DCFEM2 and the constitutive relationships that are required for such a model. The code is verified with existing analytical solutions and then is used to simulate the observed gassy soil behaviour under laboratory and field conditions.

Over the past decades simulation techniques, and in particular finite element method, have been used successfully to predict the response of systems across a whole range of industries including aerospace, automotive, chemical processes, geotechnical engineering and many others. In these numerical analyses, the behaviour of the actual material is approximated with that of an idealised material that deforms in accordance with some constitutive relationships. Therefore, the choice of an appropriate constitutive model that adequately describes the behaviour of the material plays an important role in the accuracy and reliability of the numerical predictions. During the past decades several constitutive models have been developed for various materials. In recent years, by rapid and effective developments in computational software and hardware, alternative computer aided pattern recognition techniques have been introduced to constitutive modelling of materials. The main idea behind pattern recognition systems such as neural network, fuzzy logic or genetic programming is that they learn adaptively from experience and extract various discriminants, each appropriate for its purpose. In this thesis a novel approach is presented and employed to develop constitutive models for materials in general and soils in particular based on evolutionary polynomial regression (EPR). EPR is a hybrid data mining technique that searches for symbolic structures (representing the behaviour of a system) using genetic algorithm and estimates the constant values by the least squares method. Stress-strain data from experiments are employed to train and develop EPR-based material models. The developed models are compared with some of the existing conventional constitutive material models and its advantages are highlighted. It is also shown that the developed EPR-based material models can be incorporated in finite element (FE) analysis. Different examples are used to verify the developed EPR-based FE model. The results of the EPR-FEM are compared with those of a standard FEM where conventional constitutive models are used to model the material behaviour. These results show that EPR-FEM can be successfully employed to analyse different structural and geotechnical engineering problems.

Synthetic polymer fluids have been used as an alternative to conventional bentonite slurries for the excavation of foundation elements over the last two decades. Thus far their use has excited polarised views in the industry – some construction professionals avoid using them after hearing ‘horror stories’ about past failures whereas others believe that they are the future of support fluids because of the many benefits that they can offer. This situation has been found to be mainly due to a lack of understanding of what polymers really are and their properties. To address this issue, the various polymer products currently available in the market have been categorised based on their intended functions, and the fundamental properties of some of them have been characterised. The research showed that significant difference exists between polymers even of the same chemical family. Based on a series of rheological experiments, it was found that the properties of polymer fluids are highly dependent on the governing shear rate, time, and a range of site dependent factors. The research also showed that the current procedure for the interpretation of viscosity data is incorrect. After an analysis of the properties of currently available polymers, the latter part of the thesis describes a field trial carried out at a site in Stratford, East London. The aims of the trial were twofold: to compare the performance between piles constructed using bentonite and polymer fluids, and to assess the effect of extended pile bore open time for pile bores supported by polymers. From the results, it was found that the polymer piles showed much stiffer load–settlement response than the bentonite pile, and that a pile bore open time of up to 26 h had no adverse effect on the pile performance. The load test results have been carefully back-analysed using a range of methods. It was found that, compared to their bentonite counterparts, polymer fluids can improve the interface shearing resistance with Thanet Sand and the shear modulus of the Lambeth clay.

Due to the findings of extensive damage on geogrids used in a road embankment in northern Sweden, the Swedish Transport Administration (TRV) started to investigate the reason of these damages. Since the geogrids were installed at low temperature, below 0°C, it was suspected that the damages were connected the low temperature. To analyse whether low temperatures have an influence on the extent of installation damages, both a half-scale setup and temperature controlled tensile tests have been carried out on geogrids. In total five different types of geogrids have been tested; 3 extruded polypropylene geogrids, 1 woven PET geogrid, and 1 welded PET geogrid. All geogrids had an aperture size of approximately 35 mm and specified tensile strength of approximately 40 kN/m. The Half-scale tests was conducted by building a small road embankment inside a freeze container, at the Luleå University of Technology (LTU). The embankment contained crushed aggregate, type 0-70 mm, and geogrids. The purpose of the half-scale test was to simulate installation of geogrids at different temperatures and thereby investigate whether low temperatures have an influence on the rate of installation damages. The half-scale test was done for each type of geogrid at the temperatures: +20°C, -20°C and -30°C. First, the geogrid was covered by 150 mm of crushed aggregate. Then a vibratory plate (160 kg) was used to compact the crushed aggregate. After each installation, the crushed aggregate was removed carefully by vacuum suction. The geogrid was removed and then analysed by visual control and tensile tests conducted according to ISO 10319:2008 (wide width tensile test). Results from the half-scale tests indicate that 2 out of 5 of the tested geogrids were affected by the testing procedure. The results indicate that: -        one of the geogrids of polyprophylene (here referred to as G2) was more damaged at lower temperatures compared to installation at +20° C. -        the geogrid of woven PET (here referred to as G5) was less damaged at lower temperatures compared to installation at +20° C. Results for the other geogrids are either inconsistent or shows no significant variation of the measured parameters as function of temperature. Hence, these results cannot be interpreted as damage during installation. Temperature controlled tensile tests were done by tensile testing single strands from the geogrids to failure, inside a temperature controlled chamber. The purpose of these tests was to investigate how the strength properties of the geogrids are affected by low temperature. The test was repeated 5 times for each geogrid and temperature (+20°C, 0°C, -10°C and -20°C). Force and strain was measured during the tests. The results from the temperature controlled tensile tests show that the maximum strain decreases with lower temperature for all tested geogrids. The maximum strain decreased by 16% - 49% when the temperature dropped from +20°C to -20°C. The results show that the tensile strength increases with lower temperature for all tested geogrids except for the welded PET geogrid (here referred to as G1). For G1 the tensile strength decreased by approximately 7% at a temperature drop from +20°C to -20°C. For the woven PET geogrid (G5) and the polypropylene geogrids (G2-G3) the tensile strength increased between 13%-45% at a temperature drop from +20°C to -20°C. The E-modulus increased at lower temperature for all tested geogrids. The secant E-modulus at 2% strain increased by 13%-71% at a temperature drop from +20°C to -20°C. Summarized conclusions from the tests: Strength properties changed for all tested geogrids as the temperature decreased. All tested geogrids got stiffer at lower temperatures. The magnitude of the effects is different for different geogrids. The tensile strength increased with lower temperature for all tested geogrids except for the welded PET geogrid, which got lower tensile strength at lower temperature. The half-scale test indicates that the amount of installation damages at geogrids can be dependent of the temperature at installation. However, these indications can only be seen at two out of five tested geogrids. The effect cannot be connected to a specific step in the installation procedure and cannot be explained by the results from the temperature controlled tensile tests. The results from the half-scale test have a statistically low reliability since only one installation for each temperature and geogrid type was done. The compaction equipment used during the test was small, and had low compaction energy compared to a vibratory roller compactor commonly used in construction work. With respect to the discussion above, further studies should be focusing on developing the half-scale test. It is suggested that the test is scaled up to a full-scale test in order to simulate a real installation as close as possible. The test should also be conducted several times for each geogrid at each temperature in order to enable statistical analyses.

Tailings is leftover material from mining industry and is produced in huge quantities approximately 70-99% of the ore production.  Tailings material is stored as impoundments by constructing tailings dams which are often constructed with tailings material itself. Tailings are artificial material and the mechanical behavior of tailings material upon loading is different as compared to natural soil materials. There are number of dam failures reported every year which has severe impact on inhabitants and environment nearby. Considering the failures of tailings dams and consequences there is a need to understand the tailings material in depth for safe existence of these dams. The confident dam design can assure the safe existence of tailings dams for long term as these dams are presumed to function for generations to come. The material properties in tailings dams can change during operation due to raising of new layer. Raised new layer can change stress level, which in turn may change the material properties in terms of strength, pore pressures, grain sizes etc. Today mostly tailings dam are designed by performing analysis for safety for existing and future rasings as well. These analyses are based upon a for certain factor of safety. Not very much can be done with design and analysis for tailings material if the material is not described very well. Understanding of tailings material in depth can provide help for detailed material parameters which later can be used in safety assessment for future raising and changed conditions in dam. This study presents the work carried out on tailings material from a Swedish tailings dam. The study is conducted on undisturbed and disturbed tailings material. The undisturbed tests are carried out to understand material properties as per in-situ conditions. Whereas disturbed materials are used to created different materials with different particles sizes. Initially in this study the basic properties of tailings materials are studied e.g. specific gravity, phase relationships, particle sizes, particle shapes and shear behavior on collected samples at various depths. During direct shear tests, the unexpected vertical height reductions were observed, these results are presented in this study. The comparison of strength parameters by direct shear and triaxial tests on material from various depths is also done and presented. Based on results from direct shear, triaxial and oedometer tests on uniform sized tailings material; the evaluation of primary and secondary deformations and particle breakage and effect of vertical loads is also carried out and presented. The study also includes the comparison of strength parameters for each particles size. The breakage of particles is analyzed by sieving the material after direct shear tests followed by a particle shape study. The effect of deposition on shear strength parameters is also studied by construction of samples with different angle of deposition of material. The strength parameters of uniform sized particles in triaxial tests are also evaluated and discussed.

In addition to vertical loads, the foundations of offshore structures are subjected to horizontal loads and overturning moments as a result of environmental (wind and wave) loading. The behaviour of circular footings on cohesive soil under conditions of combined vertical, horizontal and moment (V, H, M) loading is the primary concern of this thesis. A programme of physical model tests, involving combined loading of circular footings on reconstituted Speswhite kaolin, is reported. The shape of footing used is typical of the "spudcan" foundations of independent leg jack-up drilling platforms. Previous experience with combined loading of footings on sand has revealed that the observed load:displacement behaviour is best understood, and theoretically modelled, in terms of work hardening plasticity theory. The present tests on clay confirm this, and the results are interpreted to give empirical expressions for (i) the combined load yield surface in V:H:M space, and (ii) a suitable flow rule to allow prediction of the corresponding footing displacements (z, h, θ) during yielding. Extension to a complete plasticity model is achieved using theoretical stiffness factors to define elastic behaviour, and theoretical lower bound bearing capacity factors (derived specifically for this work) to define the size of the yield surface as a function of vertical penetration. The predictive capabilities of the numerical model are evaluated by retrospective simulation of various footing tests. Finally some plane frame structural analyses of a representative jack-up unit are described; some of these analyses incorporate the plasticity-based numerical model of spudcan footing behaviour under combined loads.

Given their potential applications for a number of engineering purposes, the geomechanics of shale reservoirs is becoming one of the most important issues in modern geomechanics. Borehole stability modeling, geophysics, shale oil and shale gas reservoirs, and underground storage of CO2 and nuclear waste are some of these potential applications to name a few. The growing interest in these reservoirs, as a source for hydrocarbons production, has resulted in an increasing demand for fundamental material property data. Laboratory analysis and constitutive models have shown that rock elastic and deformational properties are not single-value, well-defined parameters for a given rock. Finding suitable values for these parameters is of vital importance in many geomechanical applications. In this thesis an extensive experimental program to explore geomechanical properties of shale was developed. A series of triaxial tests were performed in order to evaluate the elasticity, yielding, and failure response of Marcellus shale specimens as a function of pressure, temperature, and bedding angle. Additional characterization includes mineralogy, porosity, and fabric. Rock samples used in this study came from three different locations and depths: one actual reservoir (~7,500 ft. deep), and two outcrops (~300 ft. and ~0 ft. deep).

Geothermal energy has become a focal point of the renewable energy revolution. Both shallow and deep types of geothermal energy have the potential to offset carbon emissions, reduce energy costs, and stimulate the economy. Before widespread geothermal exploration and exploitation can occur, both shallow and deep technologies require improvement by theoretical and experimental investigations. This thesis investigated one aspect of both shallow and deep geothermal energy technologies. First, a group of shallow geothermal energy piles was modeled numerically. The model was constructed, calibrated, and validated using available data collected from full-scale in-situ experimental energy piles. Following calibration, the model was parameterized to demonstrate the impact of construction specifications on energy pile performance and cross-sectional thermal stress distribution. The model confirmed the role of evenly spaced heat exchangers in optimal pile performance. Second, experimental methods were used to demonstrate the evolution of a fractured granite permeability as a function of mineral dissolution. Steady-state flow-through experiments were performed on artificially fractured granite cores constrained by 5 MPa pore pressure, 30 MPa confining pressure, and a 120°C temperature. Upstream pore pressures, effluent mineral concentrations, and X-Ray tomography confirmed the hypothesis that fracture asperities dissolve during the flow through experiment, resulting in fracture closure.

Early warning of slope instability will enable evacuation of vulnerable people and timely repair and maintenance of critical infrastructure. However, currently available warning systems are too expensive for wide-scale use or have technical limitations. The acoustic emission (AE) monitoring approach using active waveguides (i.e. a steel tube with granular backfill surround installed in a borehole through a slope), in conjunction with the Slope ALARMS AE measurement system, has the potential to be an affordable early warning system for slope instability. However, the challenge has been to develop strategies to interpret and quantify deformation behaviour from measured AE. The development of an approach to quantify slope deformation behaviour from measured AE will enable the AE monitoring system to provide early warning of slope instability through detecting, quantifying and communicating accelerations in slope movement. Field monitoring and full-scale physical modelling have been conducted to characterise the AE response from the system to both reactivated slope movements and first-time slope failure. Definitive field evidence has been obtained showing AE monitoring can measure slope movements and generated AE rates are proportional to slope displacement rates, which was confirmed through comparisons with both conventional inclinometer and continuous ShapeAccelArray deformation measurements. A field monitoring case study demonstrated that the AE approach can detect very slow slope movements of 0.075 mm/day. In addition, the concept of retrofitting inclinometer casings with active waveguides to convert the manually read instrument to a real-time monitoring system has been demonstrated using a field trial. Dynamic strain-controlled shear tests on active waveguide physical models demonstrated that AE monitoring can be used to quantify slope displacement rates, continuously and in real-time, with accuracy to within an order of magnitude. Large-scale first-time slope failure experiments allowed the AE response to slope failure to be characterised. AE was detected after shear deformations of less than a millimetre in previously un-sheared material, and AE rates increased proportionally with displacement rates as failure occurred. The AE rate-displacement rate relationship can be approximated as linear up to 100 mm/hour and shear surface deformations less than 10-20 mm. At greater velocities and larger deformations the gradient of the relationship progressively increases and is best represented using a polynomial. This is because complex pressure distributions develop along the active waveguide analogous to a laterally loaded pile, and the confining pressures increase. Variables that influence the AE rate-displacement rate relationship have been quantified using physical model experiments and empirical relationships. A framework has been developed to allow AE rate-displacement rate calibration relationships to be determined for any AE system installation. This provides a universal method that can be used by practitioners when installing AE systems, to calibrate them to deliver alarm statuses/warning levels that are related to slope displacement rates. Use of this framework has been demonstrated using a case study example, and decision making protocols have been suggested that use trends in alarms with time to trigger decisions, which could be to send an engineer to inspect the slope, manage traffic, or evacuate people.