Suction caissons have been used extensively for anchoring and supporting the offshore installations like oil platforms and wind turbines. These foundations are normally subjected to complex combinations of the vertical, horizontal and moment loads (i.e. V, H, M) from the self-weight, wind, wave and currents. In the past decades, extensive studies have been conducted to investigate the combined V-H-M loading behaviour of suction caissons in clay. However, most existing studies are focused on the ultimate bearing capacity, while the deflection response is more critical in foundation design for recent infrastructures like offshore wind turbines. Due to the complex load conditions, predicting the three-dimensional (3D) deflection response of the foundation is still challenging. Machine learning (ML) appears on the research horizon due to its excellent capacity of solving nonlinear problems with desired speed and accuracy. However, conventional machine learning approaches were limited in their capacity to analyze raw natural data without artificial interventions. Meanwhile, the deep learning technique (DL), as a branch of machine learning, allows a machine to be fed with raw data, automatically extract the features, and discover intricate structures in high-dimensional data. The deep learning technique has been used in many fields like language translation, auto-pilot and image recognition. And Deep neural networks, including deep learning algorithms and architectures, are gradually being developed. In light of these backgrounds, this study proposed to develop a deep learning based surrogate model to predict the 3D deflection response of suction caissons under combined V-H-M loading. The advanced three-dimensional nonlinear finite element (FE) simulations under complex V-H-M loading paths were performed on suction caissons of different geometric configurations and in clay soils with different stiffness and strength properties. The 3D FE simulation data was then used to train the deep learning based design model. Three popular neural network structures, i.e., Feed forward Neural Network (FNN), Convolution Neural Network (CNN), Recurrent Neural Network (RNN) have been employed to develop the hybrid surrogate design model. In this study, two different training strategies were proposed for this geotechnical problem. In the first category, the 3D load-deflection behaviour of suction caisson is idealized as a point-to-point mapping problem, i.e. mapping between the deflections (i.e. displacement and rotation) with loads (i..e force and moment). This task was achieved by Fully-Connected Neural Network model (FC-NN) based on FNN, One Dimension Convolution Neural Network model (1D-CNN) based on CNN and Long Short Term Memory model (LSTM) based on RNN. In the second training strategy, the load-deflection response was idealized as a time series process, a line-to-line mapping problem, mapping between the past loading paths (i.e. 10 groups of forces and moments) with future loading paths (i.e. 90 groups of forces and moments). Besides the three neural network models mentioned before, another two complex and advanced models, LSTM Model combined with convolution neural network (1D-CNN+LSTM) and Temporal Convolutional Network model (TCN), are also applied for temporal prediction. The performance and training efficiency of these models were also systematically evaluated by interpolation and extrapolation experiments. Basically, all the models can well capture the 3D deflection response of the foundation with significantly high accuracy (i.e., root mean squared error is smaller than 0.05 and coefficient of determination is near 1.000) than the traditional design approach (such as macro-elements model), and with greater efficiency than the 3D FE simulations. Among all the models, the TCN model has the highest prediction accuracy and robustness. However, the FC-NN model has the simplest model structure and highest computational efficient in learning the non-linear relationship between deflection response and V-H-M load. Besides capturing the relationship between input and output, the deep learning model can also assist to identify the intrinsic failure mechanism. By observing the fluctuation of generalisation ability, the evolution of the failure mechanism of suction caisson with embedment depth was revealed.

Monopiles with large diameter (larger than 6 m) and low aspect ratio (less than 6) are increasingly used in offshore wind farms. These foundations demonstrate a rigid response under lateral loading. The validity of the existing design methods, that are based on small diameter flexible piles, has been questioned by both the industry and researchers. In addition, the monopiles are subjected to both lateral and vertical loads. The influence of vertical load on the lateral design of short rigid monopiles in clay soil is not clear. This study aims to perform a comprehensive study on the influence of vertical load on the lateral response of monopile foundations in clay soil.

All analysis in this study was performed using 3D finite element modeling in PLAXIS 3D software. The NGI-ADP constitutive model was adopted to simulate the nonlinear mechanical behaviour of clay. Considered in the analysis is a short rigid pile with a diameter of 10 m (L/D = 3) and a long flexible pile with a diameter of 2 m (L/D = 15). The analyzed clay soil profiles consist of a normally consolidated clay soil and an overconsolidated clay soil with a constant undrained shear strength profile equal to 30 kPa. For each pile in each type of clay soil, a pure lateral loading scenario is performed first to assess the validity of current design methods. Subsequently, a combined loading scenario is performed to assess the influence of vertical loading on the lateral behaviour of rigid monopile in clay soil.

Results of the pure lateral loading scenario suggest that current design methods heavily underestimate the lateral capacity of rigid monopile foundations in both clay soil profiles analyzed. According to the findings of this study, it can be concluded that current design methods are not fit to provide an accurate assessment regarding the lateral load response of rigid monopile in clay soil. In order to correctly assess the lateral load response of rigid monopile in clay soil, a method consisting of a 3D finite element model akin to the model used in the research or a PISA design model is advised. A potential third design method, the 1D rotational spring model, is also proposed.

Results of the combined loading scenario suggest that the presence of vertical loading causes a decrease in lateral and moment capacity of the rigid pile in both clay soil profiles analyzed. However, the influence is negligible when the vertical load magnitude is smaller than 50% of its bearing capacity. To quantify the influence of vertical load on a monopile foundation, a series of load analysis were performed on a real offshore wind turbine with a 5MW power capacity. It was found that the vertical load on a typical monopile foundation in clay is around 27% of its bearing capacity. According to the findings of this study, it can be concluded that the influence of vertical load on the lateral response of rigid monopiles in clay soil is limited and can be ignored in foundation design.

The Dutch pile design method, NEN9997-1, classifies screw and screw-injection piles as fully displacing. For these pile types, the design code prescribes a base factor αp 0.63 and shaft factor αs in sand of 0.009. However, an ongoing TU Delft research programme on these pile types has indicated findings conflicting with NEN9997-1. This thesis investigates whether screw and screw-injection piles should be classified as fully displacing piles and how the design process can be improved, through the interpretation of existing load tests.
The thesis compares measurements of pile load tests to the load-settlement behaviour of fully displacing and (partly) soil replacing piles, including the effect of limiting qc to a maximum of 15 MPa for shaft friction. From appropriately instrumented tests, αp and αs factors are determined and compared to the prescribed factors. Additionally, a shear box test is performed in order to investigate debonding between the grout body and steel tube of a screw-injection pile.
Interpretation of the load tests strongly signify that the load-settlement behaviour of screw and screw-injection piles does not resemble that of fully displacing piles, but rather (partly) soil replacing piles. Determined values of αp range from 0.23 to 0.35, while values for αs in sand range from 0.011 to 0.012. Limiting qc along the shaft is shown to produce less realistic capacity and behaviour predictions when compared to measured test data. The shear box experiments indicate that in dense soils with high qc values, debonding between the grout and steel tube of a screw-injection pile under high load can occur.

Digital Soil Mapping (DSM) of soil types in geotechnical project areas is a top priority. These maps are often used in decision making and can have significant consequences related to costs and risks. Usually, these maps are generated by digital soil models that interpolate soil types at known locations. In practice, conventional spatial interpolation techniques are still often used for DSM of soil types, such as inverse distance weighting and kriging. However, conventional models are not well suited for predicting or interpolating soil types because of their inability to deal with categorical data properly. Besides, the design of the conventional models does not allow for incorporating the abundance of meaningful covariate information that is available nowadays. The flexibility of machine learning algorithms vanquish both problems and has become increasingly popular for DSM of soil properties in recent years. The results of machine learning techniques for DSM of soil properties are promising and generally outperform conventional models. However, few studies have used machine learning for DSM of soil types and is therefore still a relatively unknown field. Moreover, at the time of writing, there are no studies that use sequence models for DSM of soil properties or types. Hence, the author proposes to introduce a new method for DSM of soil types, namely a Long Short-Term Memory (LSTM) network. The intuition behind this introduction is that the spatial correlation can be captured in sequences and can improve soil type prediction.

Real project data from a cable burial project is used to evaluate and compare the performance of the conventional interpolation methods triangulation and kriging, the machine learning models random forest and XGBoost, and the newly proposed deep learning model LSTM. The project data consist of 757 vibro cores (VC), 718 cone penetration test (CPT), bathymetry data and sub-bottom profilers. The geotechnical data, i.e. VCs and CPTs, is received on separate PDF pages that require to be digitized first. This thesis describes a simple yet precise manner to extract this data from the PDFs. The VCs and CPTs are provided with a soil type interpretation and can be used directly for developing the models. The data is split into a training set to develop/train the models and a test set for evaluation. Ultimately, the best performing model is used to build a 3D stratigraphic soil model for the project area with associated prediction accuracies.

All state-of-the-art techniques outperform the conventional models and especially in predicting minority classes. The best performing model is random forest with an overall accuracy of 85.44\% and is comparable to the performance of XGBoost of 85.11\%. LSTM network achieved a slightly lower accuracy of 84.27\%. The results show that LSTM is suitable for DSM of soil types and has considerable potential for improvement as only a few possibilities of the model have been examined.

The increasing global demand of renewable and clean energy has led to the exponential growth, development, and interest in the offshore wind energy industry and expansion towards earthquake-prone areas. Offshore wind turbine structures, typically supported by a tubular monopile foundation, are increasing in size to meet the increasing human demands. In the constant ongoing debate in search of a balance between design accuracy and efficiency, general consensus is yet to be found in search of accurate yet simplified representation of soil-pile interaction. A Winkler foundation principle has shown to be a good compromise regarding this discussion but current knowledge is mainly based on (pseudo-static) small soil-strain inducing wind- and wave load-cases. Adopting this principle, the soil-structure interaction mechanism is represented by local lateral soil reactions (springs with distributed stiffness in mechanical formulation). Strong ground shaking induces shearing and volumetric variation of the soil particles. Through hysteresis the soil material exhibits energy dissipation: hysteretic damping. Accounting for hysteresis is considered computationally more demanding than when the soil continuum is assumed elastic but this is an assumption which only holds when soils undergo very small soil strains.
This thesis explores the amount of energy dissipated as a result of hysteresis during seismic response of an offshore wind turbine and the applicability of such damping in a local linear visco-elastic manner. In order to obtain insight in this nonlinear energy-dissipation mechanism associated with the hysteretic offshore wind turbine model under seismic excitation, a Python code was developed that calculates the energy dissipation of each load-cycle separately. The developed energy dissipation assessment algorithm is effective in application of arbitrary hysteretic response and unloading-reloading rules.
The hysteretic nature of the soil-pile interaction springs in question are calibrated against the widely applied API p-y, force-displacement curves. Unloading-reloading rules are specified to define the load-cycles. Boulanger et al. describes such unloading-reloading rules for pile application under seismic loading. The applicability of these backbone curves and unloading-reloading rules remains questionable in application of rigid monopile foundations. Despite not representing the accuracy of true soil-monopile interaction, obtained results in this research may support the exploration of innovative unloading-reloading rules.
The developed energy dissipation algorithm is proven to be a powerful tool in identifying the amount of energy dissipation over a total timeseries. Reasonable agreement in peak (maximum observed), Ultimate Limit State, deflection and bending moment seismic response at mudline and tower top has been found between a hysteretic supported model and equivalent elastic models using a single (load-dependent and depth-dependent) equivalent damping coefficient in parallel with each soil spring. Representing the hysteretic energy dissipation mechanism using viscous dampers with constant damping coefficients has therefore proven to be an effective modelling strategy to account for the damping mechanism of plastic unloading-reloading rules without accounting for hysteresis. The effectiveness of an equivalent elastic modelling strategy reduces when the response undergoes substantial permanent plastified displacements. A typical property which is unable to be simulated under the application of an elastic modelling strategy.

The numerical modelling of a cone penetration test (CPT) has long been a challenging task due to the large deformations associated with the penetration of a CPT. Recent developments in advanced numerical methods have shown promising results in overcoming these difficulties by using the Material Point Method (MPM). In this thesis it is researched whether the MPM is able to reliably produce CPT results in dry sand by using a state-dependent constitutive model. Calibration chamber (CC) tests are modelled for dry sand and results are compared with experimentally performed CC tests in the laboratory. Features regarding the numerical setup and applied boundary conditions which quantitatively influence modelling results are identified and assessed before the model is validated to real CC test data. Validation results show that the model is able to accurately produce cone resistance values for different types of sand for soil states that can be categorised as moderately-dense to dense. Last, it is shown how parameters within the constitutive framework affect the model output and a quantification of the sensitivity of the parameters to model results is presented.

Offshore wind power structures are subjected to cyclic loading. Several loads like wind and wave loading are acting on the structure and therefore its foundation. It is important to involve the impact of this cyclic loading into the design of the structure. A cyclic load influences the strength and deformation characteristics of the soil. Due to this the cyclic loading the cyclic shear strength will decrease. This cyclic shear strength depends on several factors which can be influenced by soil properties and shear mechanisms. The aim of this Master thesis is to provide correlations between index parameters and the cyclic simple shear strength to observe how and which parameters influence the shear strength behaviour due to cyclic loading. This research focuses on the results from cyclic simple shear tests and only for cohesionless soil from North Sea sand. The project material was obtained from several offshore windfarm projects by Fugro, namely The Hollandse Kust West ,The Hollandse Kust Noord and The Hollandse Kust Zuid. Available data from these project locations is analysed and translated to cyclic resistance curves. In these curves the number of cycles to reach a nominal strain failure is plotted against a normalised cyclic shear strength, also called a cyclic stress ratio. The strain failure criterion was chosen at 3.75\% and a commonly equivalent number of cycles of 10. The soil parameters that are elaborated for this research are based on characteristics compared to density of the soil, grain structure and in-situ characteristics. These parameters are the relative density, fines content, mean particle size and the normalised cone penetration resistance. The relative density is an unknown parameter in this research and is determined in two ways. First, by considering in-situ data where it was measured along depth and an assumption was made about the real value based on the known depth range of the borehole sample. The second method was based on the initial void ratio measured in the laboratory to obtain a relation between the in-situ measured relative density and the initial void ratio, which is called the theoretical value of the relative density. From this thesis, it was concluded that due to the small range of data it was not possible to provide clear correlations and to observe trends of soil index parameters with the cyclic shear strength. By looking at the total range of the cyclic response of the data in the cyclic resistance curves and the elaborated soil parameters, the relative density and the fines content influence the shape of the power failure lines. A relatively high value of the relative density results in a larger range of different cyclic stress ratio values, i.e. a failure line with a higher gradient. An increment in fines content results in a relatively lower values for cyclic stress ratio and do not differ much by an increment of load cycles. This observation is based on the available results from this research with a relatively low amount of data and a small range of fines content.

Six cyclic settlement models for sand are evaluated to analyse the settlement of automatic stacking crane (ASC) rail tracks at the Rotterdam World Gateway (RWG) container terminal. During Phase 1 of the RWG container terminal settlement of the rail tracks occurred at multiple locations after the ASCs became operational. This has repeatedly led to (unplanned) downtime of parts of the RWG container terminal due to rail track maintenance. Settlements are caused by densification of the sand fill, which is a result of the cyclic load applied by ASCs moving continuously over their rail tracks.

 

The aim of this research is to contribute to prevent unplanned downtime in Phase 2 of the RWG container terminal due to rail track settlements. Also, reliable settlement predictions can be used to determine the intensity and extent of the ground compaction that are needed to meet the settlement requirement of 20 mm for ASC rail tracks.

 

The cyclic settlement models, which have been validated to predict the cyclic settlement of rail tracks and shallow foundations, are obtained from literature. The available soil data include CPT’s, boreholes and standard laboratory soil testing. In addition, settlements of the ASC rail tracks in Phase 1 had been measured for a period of almost one year. The cyclic settlement models are evaluated at six different locations, where the sand is medium to very dense and settlements up to 32 mm have been measured. The load is modelled as a quasi­static load equivalent to a vertical stress of 60 to 90 kPa applied to the ballast-­sand interface. The model parameters of the cyclic settlement models are determined by correlation, (FE) modelling of the first load cycle, extrapolation and estimation.

 

The zone of influence was found to reach around 6 m below the shallow foundation. Densification of the sand fill is substantial within the entire zone of influence. The maximum densification was found not to coincide with the minimum void ratio, it is a variable that depends on the initial state of the sand and the loading and soil conditions. After order 10⁴ load cycles densification of the sand was found to become negligible. To meet the settlement requirement for ASC rail tracks the sand fill must consist of sand layers with a minimum and average relative density of at least 65% and 85%, respectively.

 

Cyclic settlement increases with the number of load cycles, amplitude of the load and extent of the zone of influence and decreases with relative density, stiffness of the sand and volumetric threshold strain. However, correlations used to calibrate the model parameters lead to model predictions that are over- or insensitive to parameters that affect the cyclic settlement. The cyclic settlement predictions of the terminal density model are most reliable and match best with the settlement measurements, for loose and medium dense sand the model predictions underestimate the settlement.

 

Instead of using correlations to obtain the model parameter values and decrease their uncertainty it is recommended to measure the:

·         disturbance of the sand fill underneath the ASC rail tracks due to construction;

·         maximum densification of the sand underneath ASC rail tracks in Phase 1 at locations where rail track settlement has stopped, i.e. where the sand reached its maximum densification;

·         model parameters that characterise the cyclic densification behaviour of sand in cyclic soil tests.

This will improve the reliability of the cyclic settlement predictions of ASC rail tracks constructed on a sand fill. To validate the cyclic settlement models for ASC rail tracks on sand, measurements of the settlement with depth as function of the number of load cycles are needed.

In this master thesis the behaviour of a high speed railway track constructed on top of a pile foundation with settlements under a dynamic load has been analysed. Therefore, a model was developed in the software program DARTS, dynamic analysis of rail track structures. In DARTS the railway track was modelled by stiff and elastic layers, which were loaded by a moving train load. The train was modelled as a mass spring system. The results of DARTS were analysed and validated with field measurements of the HSL. After the validation of the model multiple simulations were performed to analyse the structure for different uncertainties and possible solutions. These simulations showed that the model in DARTS is sensitive for changes of the pile stiffness. The results also show that it is difficult to model permanent displacements in DARTS.

The material point method (MPM) is gaining increasing amounts of attention due to its capacity to solve geotechnical problems involving large-deformations. While some problems require dynamic analysis, simulating the (infinite) continuous domain using typical Dirichlet (fixed) boundary conditions induce spurious reflections causing (1) unrealistic stress increments at the domain oundary and (2) the appearance of multiple unnatural stress waves in the domain. Aiming to eliminate this numerical artifact in MPM, two solutions for absorbing boundary conditions found in FEM are implemented and investigated; these are (1) a viscous boundary condition and (2) a viscoelastic boundary condition. The use of such dynamic boundary conditions in MPM is scarce and no validation of them has yet been presented in the literature. In this work, these absorbing conditions are implemented alongside other recent developments, which improves the numerical stability (Double-Mapping, Generalized Material Point Method, Composite Material Point Method), using two approaches: (1) directly imposing at the external active boundary nodes and (2) imposing via shape function interpolation. The proposed solutions are then validated with a one-dimensional benchmark: a soil column under dynamic load in small-deformation and large-deformation, and a 2D symmetric plane strain model under one loading pulse. The benchmark results demonstrate that the numerical reflections that lead to inaccuracies of stress and velocity can be removed using GIMP interpolation and, together with the other numerical technics, render high-quality and realistic results. A study of shallow foundation failure under repeated loading is also presented, showing the potential of applications of the proposed solution for modelling extreme geotechnical events.

In modern piling technology screw piles are used as a type of deep foundation for engineering structures, with the principal benefit of using said piles is that they offer an installation method that is virtually noise and vibration free. This makes these piles ideal for construction works in urban areas where surrounding structures can be affected by vibrations that would be produced from the installation of a driven pile. Since their initial production in the 1980s in Europe, multiple variations of screw piles are on the market today. This Msc thesis focuses on the Screw-Injection piles (SI-piles) commonly known as Fundex piles. SI-piles are a type of partial ground displacement piles where only a portion of the soil surrounding the pile is being pushed radially outwards during the rotational motion of screwing in the installation process. The other portion is transported back with grout flowing to the surface.

The current Dutch practice already have NEN guidelines on how to predict bearing capacity for SI-piles. These guidelines consist of CPT-based methods with an empirical correlation factor, the $\alpha$ pile class factor, which helps to relate the bearing capacity to the soil surrounding the pile. Nevertheless, one aspect that is not well understood is the effect that different properties of the injected grout have at the soil-pile interface and for bearing capacity. In this thesis two grout properties are being manipulated which are the Water/Binder and W/C ratios of the grout mixture, and the injection flow rate of the grout with the purpose to see whether and/or to what extent a difference exists in the shaft bearing capacity for SI-piles.

A full-scale experiment was conducted on 15 piles in order to evaluate the effect of these varying parameters. This research is composed of four targeted variations of W/B ratio and two injection flow rates, 5 groups of 3 piles each to be more precise. The piles were subjected to a static pile load test in tension, which means that the bearing capacity is composed of mainly the shaft resistance of the pile. The analysis breaks down in four main parts to analyse the indirect relationships between the properties that are accounted for in the empirical parameter $\alpha$. These four parts include the assessment of the load-displacement behaviour of the SI-piles, assessment of radial soil stress (CPT data), assessment of the records during the installation process (torque, RPM), the grout properties during installation and after 28 and 56 days of curing, and lastly, the pile shape (volume) after extraction of pile.

The assessment of the load-displacement behaviour showed that the predictions using the NEN guidelines for bearing capacity were extremely accurate for most pile groups (above 0.970 measured/predicted ratio). But for the pile groups with higher W/B ratio and with the highest flow rate (Groups C and D respectively) the measured shaft capacity would be much lower. A direct relationship between the W/C and W/B ratio is difficult to conclude since for pile B2 and C1 that had the same W/C ratio, the difference in the measured/predicted ratio was about 21\%. In the case of flow rate it is entirely seen that a higher flow rate leads to a significant decrease in measured shaft capacity. The NEN suggests a value of $\alpha_t$ for SI-piles of 0.009, yet the shaft capacity for groups with a higher W/C ratio and flow rate could be better predicted with an $\alpha_t$ $\approx$ 0.00793. Additionally, another important research objective is to try to optimise the $\alpha_t$ parameter by comparing the $q_c$ values for the pre-installation, the average post-installation and minimum value of the post-installation CPTs. This resulted in the $\alpha_t$ derived from the pre-installation CPT to have a much lower Coefficient of Variation, CoV, of approximately 0.08 whereas the average and minimum post-CPT $\alpha_t$ had a CoV of 0.12 and 0.11 respectively.

The assessment of the soil stresses is comprised of an analysis of the changes in cone resistance, $q_c$, throughout the field. The analysed data collected shows that for varying W/B ratios there is no solid relationship that relates the change in $q_c$ after the grout installation. However, a higher flow rate seems to have a significant impact on the cone resistance, leading to a general decrease of $q_c$ after installation, having a decrease as low as -16.53\% for pile D1, whereas for all other pile groups there was an increase in $q_c$ after installation, increases as high as 30\% (pile A3).

The assessment of the records during installation include the analysis of the torque during the installation process. It is seen that in both cases, high W/B ratio and high flow rate, there is a decrease in torque, but the flow rate of 115 [l/min] had a more significant impact than the increase in W/B ratio.\\
% Moreover, the 2D interpolation analysis aimed to see how post-installation CPT data should be considered. Pre-installation CPT data is sufficient to make a prediction on bearing capacity, but in this analysis both situations are being compared. This comparison resulted in that the difference between the two is minimal, the maximum difference found was in the order of $\pm$5 MPa.

The assessment of backflow grout resulted in higher W/C ratios having higher increases in density of the backflow fluid, and that high flow rate leads to a lower backflow density, this was supplemented with the sand transport data which suggests that higher W/B ratios lead to more sand transport out of the soil body. Furthermore, a inversely proportional relationship was found between W/B and W/C ratios and both the axial and bending stresses; the same inverse relationship is found with the flow rate. Additionally, the shear stress of the grout and of the soil were compared in order to determine if the failure is purely geotechnical or also structural.

The pile shape assessment resulted in a higher W/B ratio leading to a higher pile diameter, regardless of the flow rate during injection. There is also a very clear, almost perfectly linear, relationship between the mean diameter of the extracted pile and the measured shaft capacity. However not all piles were extracted and this includes piles installed with the highest W/C ratios (group C) and thus the aforementioned relationship has only been shown for a limited set of piles.

The coefficient of lateral earth pressure at rest (K0) is an important parameter in any geotechnical problem since it provides information on the initial stress state, which governs the response of the soil to the proceeding stress changes. A proper determination of the stresses in any geotechnical problem is required to be able to predict the soil behaviour. Any change in the conditions the soil is subject to can alter the value of K0, how this value changes for conventional stress histories such as loading and unloading is defined well. However, the influence of unconventional stress histories such as creep or drying/wetting cycles is not thoroughly understood. Since every soil is subjected to natural environmental stresses, resulting in creep and unsaturated conditions, it is interesting to look at their influences of the soil stress state expressed by K0.

The goal of this work is to determine if it is possible to better understand what is happening to K0 during creep and under unsaturated conditions and if the prediction of K0 can be improved by accounting for these phenomena, with the focus being on clays. Literature showed that for saturated samples, the value of K0 increases with time during creep. For unsaturated conditions it was found that K0 decreases with an increase of suction.

In order to see if it is possible to improve the prediction of K0, a model needed to be constructed. The starting point of this model was a saturated, elastoplastic model based on the SANICLAY model. The first step in extending this model was to include viscosity which was done by adopting Perzyna’s overstress approach. The model was validated to experimental data on OostVaardersPlassen (OVP) clay obtained from literature and the validation showed that the model was satisfactory in predicting the soil behaviour. Accounting for unsaturated conditions was done by adopting the average soil skeleton approach. Implementation was initially done in the original elastoplastic model. Again, the model was validated to experimental data obtained from the literature, this time unsaturated loading/unloading tests on London clay (LC) were used. The results showed that the model prediction was accurate up to suctions up to 600 kPa. The final step in the model development was to include both Perzyna’s and the average soil skeleton stress approach in the basic model giving an unsaturated elasto viscoplastic model version. Unsaturated creep tests on London clay were used to validate the model but the results showed that an uncoupled stress-suction approach gave inaccurate predictions. The viscous nucleus in Perzyna’s approach was changed to become suction dependent and the results showed that the experimental data could be reproduced reasonably well.

The unsaturated elasto viscoplastic model was then used to analyse K0 during the unsaturated creep tests. The results showed that the model predicted a decrease in K0 with time for low loads and high suctions. For higher loads and low to moderate suctions, the model predicted an initial increase followed by a decrease. For all cases it was found that the value of K0 decreased with suction. The role of anisotropy on the model prediction was analysed by predicting the change in K0 using an isotropic version of the model. This version also predicted a decrease at low loads and high suctions but an initial increase was no longer predicted to decrease. The decrease of K0 with suction was still observed. No experimental data was available to confirm either of the findings, comparing the results with the literature study showed that the decrease of K0 with suction was previously observed. The decrease with time on the other hand was not found in previous work. However, the creep tests in previous works were performed on saturated samples and in general for a shorter time period which could show different results.

It is concluded that by accounting for creep and unsaturated conditions, the qualitative prediction of the soil behaviour can be improved. By accounting for coupled behaviour through an unsaturated viscous nucleus, the currently available unsaturated and time dependent experimental deformation data can be simulated accurately. The prediction of the change in stress state, due to these natural phenomena, is likely to be improved as well since the outcome of the model matches findings from the literature. However, due to the lack of experimental unsaturated time dependent data stating the change in K0, no conclusions can be drawn on the importance of anisotropy and the quantitative model performance. What this work does offer is a good modelling tool to support future experiments or investigations into unsaturated creep behaviour.

This research looks at the workings of three design methods for piles loaded in compression and tension by modelling capacity for different soil profiles, varying parameters and load tests. Compared methods are from the NEN 9997-1 (NEN), CUR 2001-8 report (CUR) and a newly developed international standard (ISO). It is generally accepted that the NEN is overly simplistic and research shows the CUR to be generally over-predicting capacity, but the implications of these flaws are not in detail investigated. Per method, the behaviour of shaft and base resistances/capacities and total axial capacity were modelled for constant cone resistance, real CPT's and applied to case studies. Case studies, where possible, are focused on Dutch soil condition to see how the ISO would apply. The capacities were modelled over the whole domain of the CPT to get insight in general behaviour of the methods and to see where potential weaknesses lie. Also an evaluation on installation effects as residual stresses and pile ageing was done.

The CUR and NEN both have demonstrable flaws that negatively impact their capacity predictions, and most notable are their approach on shaft friction that translates to non-optimal shaft capacity profiles. Also, their response on increasing diameters of concern, as design methods are in enormous disagreement on capacities for piles with diameters larger than 1 m. Predictions compared to each other can vary threefold and differ with 8.5 MN.

The CUR generally over-predicts capacities in looser sands but under-predicts for deeper tests. The NEN generally under-predicts for denser sands but is sometimes largest of three for looser sands. General under-prediction does not necessarily mean the method is conservative. The ISO performs in most cases best, and uses a predictive plug length to influence both shaft and base capacity. The NEN is poorly suited for design, the CUR is in particular unsuited for longer piles or diameters larger than 1 m. The ISO is an improvement on current design guidelines for the Netherlands and a good step in incorporating plugging in the design of open-ended piles.

The current Dutch norms for the macro-stability of ood protection embankments and more specifically the determination of SHANSEP parameters has been subject of debate since its implementation in 2017. The main concerns arise from the apparent over-conservative nature of the guidelines and the apparent underestimation of shear strength of organic and silty clays. In this Master thesis, an assessment of the normalised normally consolidated undrained shear strength is performed for a dike stability reinforcement project between Gorinchem and Waardenburg in the South west of the Netherlands on a problematic clay layer called the `Gorinchem Clay' throughout this thesis. The main goal is the conrmation of the current norms or towards a more optimised and less conservative assessment of the strength parameters and in doing so, discusses one of the most recurrent questions in geotechnical engineering: do some of the measurements from laboratory tests reveal true material behaviour or do the limitations of such laboratory tests produce this particular behaviour? This analysis was performed on Triaxial Compression, Direct Simple Shear and Triaxial Extension tests from which the SHANSEP parameters are derived as input for the available shear strength along a slip surface according to the methodology developed by Ladd (1991) and following the Critical State Soil Mechanics. The structure of this thesis is as follows: rst, a literature study is performed on the strength parameters assessment in chapter 1, the principles of the Critical State Soil Mechanics and the diculties encountered in the determination of the strength parameters from laboratory tests in chapter 2. The laboratory results are then exploited in chapters 4 to 5 according to the current guidelines and more extended methods in which the limitations of the Classical Critical State Soil Mechanics are shown. The next chapters focus on a more fundamental understanding of the considered material consisting of modelling the material behaviour in chapter 6 using a simplied academic constitutive model featuring non-associative elasto-plasticity with mixed volumetric and deviatoric hardening allowing for hardening or softening. The outcomes of this thesis show Critical State conditions as traditionally understood could not be reached reliably for undrained conditions in Triaxial Compression, Triaxial Extension and Direct Simple Shear. The tests showed to be particularly unreliable beyond 10% axial strain in Triaxial Compression and Extension, and beyond 15% shear strain in Direct Simple Shear. The diculties encountered in determining the Critical State friction angle and undrained shear strength were shown to be minimised by performing drained and undrained Triaxial compression tests on slightly over-consolidated soil samples. Additionally, the limitations of the Classical Critical State theory were highlighted and a more advanced constitutive model including a non-associative ow rule and deviatoric hardening was successfully used to predict the behaviour in Triaxial Compression for undrained conditions. The predictions for drained conditions and particularly undrained Triaxial Extension were however limited.

The suction pile foundation is a large steel cylinder with an open end and sealed top. This foundation type is widely used in the oil&gas industry and wind energy. Experimental investigation and numerical investigation are the main two methods to understand the performance of the suction bucket foundation. The experimental studies are important and basal for design, but it is time-consuming and costly compared to numerical studies. However, the accuracy of the finite element method(FEM) in geotechnical problems highly depends on whether the soil constitutive models can correctly predict soil behaviour. In our study, we investigate four constitutive models: the Mohr-Coulomb model, the hardening soil model, the NGI-ADP model, and the hypoplastic model. To some extent, advanced soil models can better present soil behaviour than traditional ones. However, it needs more laboratory test data to calibrate the advanced model parameters. The Mohr-Coulomb(MC) model is the most common soil model, which is an isotropic model include few parameters. The hardening soil model exceeds MC model by introducing the stress-dependent stiffness and distinguishing between loading and reload- ing. The NGI-ADP model is mainly used for undrained analysis, and it is an anisotropic model which can exact match with undrained shear strength and stiffness for various failure surfaces. The hypoplastic model has no distinguishing between elastic and plastic strain. It is an inelastic(dissipative) and incrementally nonlinear soil model without the requirement of a yield surface. A soil investigation report of Block 17 offshore Angola is used to calibrate the aforementioned constitutive models. The cone penetration test(CPT), ball penetration test(BPT) and a series of laboratory tests(i.e. direct simple shear, triaxial, and oedometric tests) are exacted from the soil report and interpreted for calibration. Parameter determination procedures for constitutive models are explained. Subsequently, the consistency of the parameter set is validated by numerical simulation of direct shear and triaxial tests. The numerical experiments for a suction pile foundation whose out diameter equals four and aspect ratio equals three are carried out. Four loading cases(i.e. horizontal, vertical tension, vertical compression and vertical-horizontal-moment (VHM) combining loadings) are included in the finite element analysis. The suction pile performance(i.e. deformation and capacity) are compared among the using of different constitutive models. Additionally, the compliance matrices of the suction pile for different soil models are obtained for the structural engineer. The analyses indicated that NGI-ADP model could be the best choice for undrained analysis of suction pile foundation. This model has a robust calibration process, and well simulate the anisotropic strain-stress relationship. Additionally, the finite element results are conservative when modelling by NGI-ADP model. However, this model can not predict the right pore pressure build-up and stress-path, which may be improved by using a well-calibrated hypoplastic model.

For preparing standardized sand specimens for physical modelling in geotechnics, especially for the geo-centrifuge tests, a “line-style” sand pluviator has been recently developed by TU Delft. Controlling the falling height of sand hopper, the width of hopper’s bottom gap and the relative moving speed between the hopper and the sample box, specimens with bulk relative density ranging from 50% to 100% can be prepared by this automated machine using the coarse Merwede River sand. Besides, the periodic variation of local relative density along depth was observed using the macro-CT scanner and the features of the fabrics were investigated using the micro-CT scanner. It was also proved by a set of shallow foundation modelling tests that the sand specimens having the same bulk relative density but different heterogeneity and fabric features behaved significantly differently and further research works are recommended to explore the influences of these differences. Additionally, a partially substantiated hypothesis was proposed to conclude the general rules of the sand pluviation process, and the reliability of this hypothesis has been proved by a series of tests on the Geba sand.

Seismic design codes are currently moving from a force-based design approach to a performance-based design approach. For example, in a performance-based design approach it could be specified how many lanes must be available during the lifetime of a bridge given a certain earthquake intensity.The problem with this approach is that it is not specified what the probability must be that the performance criterion is satisfied. This raises the question whether the design codes are acceptably safe or not. Focus is laid on the Canadian Highway Bridge Design Code (CHBDC), in which a total resistance factor approach is used. Because the total resistance factor in the CHBDC is a multiplicative factor, lower resistance factors lead to stronger foundation designs. The goal of this thesis is to calibrate the design procedure in the CHBDC for geotechnical systems under seismic loading, by finding a relationship between resistance factors and the lifetime probabilities of failure of said systems. The resistance factor can then be fine-tuned to a lifetime probability of failure that is consistent with the lifetime probability of failure targeted in static design. As an example problem, the bearing capacity of a shallow foundation on a clay with a pseudo-dynamic earthquake load is tested. The research question that is answered in this thesis is: "What should the resistance factors for geotechnical seismic design be in order to achieve a target lifetime probability of failure that is consistent with static design targets?'' Not every possible combination of soil strengths and forces on the superstructure can be taken into account, and therefore the random finite element method is used in a Monte Carlo simulation. Thousands of realization sare performed for each resistance factor, design return period, and "actual''return period that the designed foundations are tested against. By seeing how many realizations of the Monte Carlo simulation fail given a certain earthquake intensity, the conditional probability of failure given that earthquake intensity can be estimated. The total lifetime probability of failure can then be estimated from the conditional probabilities of failure with the total probability theorem. As part of a parametric study, the lifetime probabilities of failure are estimated for six different scenarios, each of which has different sources of uncertainty.  The resulting lifetime probabilities of failure are interpolated in order to find a resistance factor that targets a lifetime probability of failure consistent with static design targets. Currently, the resistance factor that the CHBDC recommends for geotechnical systems under seismic loading are defined as the static resistance factor for that geotechnical system incremented with 0.20, meaning that compared to static design, weaker foundations are designed for seismic load cases. The resistance factor found in this thesis is closer to the resistance factor for static design than to the resistance factor for seismic design. It should therefore be considered to lower the seismic resistance factor to the value of the static resistance factor so that a sufficient lifetime reliability can be targeted.

In this work the installation process of a cone penetrating into a sand system is modelled using a 2D-axisymmetric material point method (MPM) model in order to investigate the installation processes and to predict the plastic radius. One meter of cone penetration is modelled under a surcharge of 50 kPa. The used constitutive model is Mohr-Coulomb Strain Softening (MCSS). The validation of the model has been conducted by comparing the MPM results to those obtained from laboratory experiments performed by Arshad et al. (2014). It was found that the qualitative behaviour from the MPM model and the experiments were in agreement but the quantitative behaviour showed some discrepancies. These discrepancies could be caused by the fact that, in the MCSS model, the stiffness is constant and independent of stress which it is not in reality. The performed sensitivity analysis showed that the stiffness has the most influence on the numerical predictions further adding to the importance of accurately capturing the soil stiffness throughout the simulation. The 2D-axisymmetric MPM model showed to have the potential to be a good tool for modelling the installation process of a cone and predicting the plastic radius.

Monopiles are commonly used as foundations for offshore wind turbine generators (WTGs). Due to the rapid growth of the offshore wind energy sector, there is increasing demand for WTGs of larger capacities which evidently leads to demand for monopiles with larger diameter. An industry standard approach for assessing pile lateral response is the p-y method; however, this method was initially developed and empirically validated for long slender piles and thus its applicability to large diameter monopiles is doubtful.

The joint academia-industry project, Pile Soil Analysis (PISA) project resulted in an improved understanding of the lateral loading response of large diameter monopiles. Based on pile load test (PLT) data and numerical modelling, a method was developed to derive all soil reaction components from advanced finite element method (FEM) calculations to be used in a one-dimensional (1D) design framework.

Cone penetration test (CPT) based approaches have been shown to provide excellent predictions for the response of laterally loaded flexible piles where the p-y response dominates. In this thesis an approach to determine the additional components of the soil reaction curves for rigid monopiles, namely the side and base shear and base moment directly from the CPT is proposed. The results are compared to soil reaction curves are extracted from 3D FEM models, and compared to field tests on monopiles in sand

Offshore wind farms are being developed at locations with moderate ice conditions such as the Baltic Sea, where drifting sea ice upon interacting with offshore structures could lead to the development of a phenomenon known as Ice-induced vibrations (IIV). These vibrations are especially severe when the turbine is idling. Current mitigation measures consist of an expensive solution of ice cones, which are only favourable when ice occurs seasonally. The main objective of this study is to investigate numerically a novel approach to mitigate the ice-induced vibrations of offshore wind turbines by means of control idling.
Three regimes of IIV are generally distinguished, viz. intermittent crushing (ICR), frequency lock-in (FLI), and continuous brittle crushing (CBR). Among these regimes, the ICR and FLI can cause significant vibrations in the offshore structure. Preceding the ice action, the rotor aerodynamics during the parked condition shows that for the wind speeds below the cut-in wind speed of the rotor, the turbine operates in the unsteady aerodynamics termed as dynamic inflow. The comparative analysis is made between the two cases: one with the ice action only, and, the other with the combined effect of ice and wind, where the rotational rotor speeds chosen are 3.0rpm, 6.9rpm and 12.1rpm. In the ice-action case, it is found that the structural response frequency during the ICR and FLI is around the first and the second natural frequency of the structure, respectively. In the case of ice and wind, it is found that the unsteady BEM method has certain limitations, especially in the ICR regime. Also, the aerodynamic damping has no notable effect on the range of IIV regimes for the rotor speed of 6.9rpm and 12.1rpm. However, it does have a significant effect for the rotor speed of 3.0rpm. The quantitative comparison of fatigue damage between the two cases showcases that for the majority of ice-sheet velocities during ICR and FLI, the damage is found to be greater in the ice and wind case. Based on the results, it is concluded that the rotor aerodynamics does help in damping the vibrations in the ICR regime, but in the FLI regime, it has no significant impact when specific ice-drift speeds are considered. It can also be confirmed that by the careful selection of the rotational rotor speed, the range of IIV regime can be influenced. However, to draw the general conclusion, the analysis needs to be conducted for varied ranges of rotor speeds. Also, the present framework of the aerodynamic model needs to be improved to capture the vortex-ring flow state to predict the rotor aerodynamics accurately for all the ice-sheet velocities.