Large-diameter driven piles are widely used as foundations for offshore wind turbines and platforms. During installation, unexpected pile running can result in rapid, uncontrolled penetration of the pile into the seabed due to the large self-weight. This paper assesses the risks of pile-running for large diameter driven piles in spatially variable soils using a procedure based on the American Petroleum Institute bearing capacity guidelines combined with Newton’s Second Law of Motion. This differs from conventional static bearing capacity analyses by calculating the pile velocity with depth. One-dimensional random fields are implemented to simulate downward spatially variable shear strengths. While fluctuations in shear strengths do not provide sizable impacts to pile shaft resistance at larger depths, end-bearing resistance variability can provide appreciable changes to pile running velocities and penetration depths. Results of the spatially variable formulation are compared with Large Deformation Finite Element simulation, providing strong agreement. Of particular importance is the conclusion that pile running analyses where soil variability is ignored can lead to unconservative estimates of velocity and depth profiles. The simplified probabilistic method for assessing heterogeneous soil properties is especially important, given that unexpected weak layers are one of the primary factors contributing to pile running.

Driven pipe piles are used extensively in coastal and offshore projects. Traditionally piles with diameters of 2–3 m were common in the offshore wind industry, however the diameter of monopiles to support a 10 MW wind turbine is more commonly 10 m. Offshore wind projects are being developed at sites with very low seabed strengths and pipe pile weights are increasing significantly. Self-weight penetration occurs when the pile is first placed on the seabed. A combination of low strength seabed conditions and increased pile self-weight leads to the risk of pile run (uncontrolled self-weight penetration) during installation at some sites. Predicting pile run risk, run velocities and penetration depths is challenging due to inherent rate effects and the large strains involved. While rapid penetration processes can be considered using both analytic methods and Large Deformation Finite Element simulations, the role of soil rigidity is seldom taken into account, despite known implications from static pile assessments. This study uses large deformation simulation with the Coupled Eulerian Lagrangian method to simulate the pile running process for five well-studied fine-grained soils with varying elastic stiffnesses. Results are compared with analytic methods, highlighting the limitations of current predictive techniques in terms of both the end tip and shaft resistance. As a corollary, a linear trend for the final penetration depth with respect to the logarithm of the soil rigidity index is incorporated in an existing analytic code based on results obtained from large deformation simulations.

Accurate determination of the load-penetration behaviour of spudcan footings is vital in predicting the performance of offshore structures. Often, insufficient site investigation data is a limiting factor in assessing spudcan penetration, with the selection of suitable parameters typically derived from triaxial tests or Cone Penetrometer Test (CPT) data. A method is presented for simulating spudcan insertion in loose to medium-dense sand, based on minimal data, whereby the CPT tip resistance is used in combination with known correlations of soil properties, allowing for back analysis of CPT profiles through Large Deformation Finite Element (LDFE) simulation. The use of LDFE allows for failure mechanisms to be determined a priori, as a result of the simulation process, while a range of constitutive models can be implemented as necessary. Friction angle, dilation angle and relative density correlations are combined with a calibration of the elastic modulus, used to develop numerical models based on a set of soil sub-layers, each 1 m in depth. The resulting soil characterisations were used to calculate the load-penetration behaviour of several spudcan geometries (which deviate from conventional spudcan shapes) using the LDFE, to assess their performance. Results generated using the proposed technique show strong agreement with a presented field observations.

Offshore jack-up rigs are most commonly founded on large-diameter conical “spudcan” foundations, which are frequently designed using traditional analytical methods for shallow footings. This paper presents the design of a spudcan installed off the coast of Tunisia. The maximum penetration depth of the footing under the available preload is predicted by a combination of analytical techniques, 2-dimensional axisymmetric modelling and 3-dimensional Finite Element Methods (FEM) using large strain arbitrary Lagrangian-Eulerian (ALE) techniques. Spudcan penetration based on FEM simulation of CPT soil profiles forms the basis of a comparison with results from the Society of Naval Architects and Marine Engineers (SNAME) guidelines. Particular attention is given to model calibration using the limited site investigation data available. Results are presented for the effect of penetrating footings on the behaviour of neighbouring footings, showing good agreement with conventional prediction methods.

This paper presents the results of compression and tension load tests performed on a single helical pile installed in dense sand. The pile was instrumented using strain gauges that allowed the shaft and base load resistance to be separated and the distribution of shaft resistance along the pile during the test to be determined. The pile was loaded first in compression, with a maintained load test, followed by a constant rate of penetration load test being performed to assess the effects of creep on the pile’s response to compression loading. The pile was then loaded in tension using a maintained load test procedure. Finite element analyses were performed using Abaqus and these helped to provide additional insights to explain the response of the instrumented pile during loading. The test showed that during compression loading, substantial bearing pressures developed beneath the pile helix, which provided the majority of axial load resistance. During tension loading, uplift pressure mobilized on the helix again provided the majority of axial resistance. The strain gauges suggested that the pile load response to compression loading was ductile. During tension loading, the pile response was brittle. Whilst load tests performed on only one instrumented pile test are presented, the use of instrumentation and finite element analyses allowed important insights into the load–displacement response of helical piles.

This paper presents the results of full-scale load tests performed to investigate the end bearing pressure mobilised by continuous flight auger piles installed in sand. In particular, the tests considered the effects of the footing width and, by varying the load test procedure (from maintained load test to constant rate of penetration), allowed quantification of creep effects. By comparing the load test results with in situ test results from cone penetration tests, correlations between the end bearing pressure mobilised at normalised settlement levels of 10% of the footing width and the cone penetration test q_c value were studied. For the maintained load tests, these correlations were found to be similar to those used in routine design practice. When creep effects were reduced using constant rate of penetration load testing, the end bearing pressure mobilised was significantly higher than that assumed in normal practice, and it was in keeping with the results of finite-element analyses performed using a soil model that ignored creep. In the final section, the field test results are compared to database pile load tests performed on non-displacement piles in sand.

A wide range of correlations have been proposed between cone penetration test end resistance, q_c, and the ultimate end bearing pressure developed by bored piles. The q_c value is typically related to the pile end bearing resistance at a normalised pile displacement equal to 10% of the pile diameter, q_b0·1, through a correlation factor, α. While it is generally accepted that constant α factors can be applied for the design of displacement piles, a combination of field tests and finite-element analyses of bored piles in sand has resulted in a myriad of design approaches, some of which suggest that α varies with pile geometry, stress level and sand density. In this paper the results of finite-element analyses are presented which suggest that a constant α factor can be adopted for the design of deep foundations in sand.

A large portion of the Irish railway network contains earthworks that were poorly constructed mainly because they were built over 100 years ago when scientific understanding of soil behaviour was not nearly as advanced as it is today. Also, engineers involved in the construction of these structures had to work with often unsuitable local materials as sound bulk fill was not able to be transported significant distances. In order to improve the information obtained from site investigation of these poorly constructed earthworks a number of authors have recently suggested complimenting direct geotechnical investigations with a non-invasive geophysical assessment. In addition to providing a review of the factors which cause failure of earthworks, this paper also discusses the capabilities of three geophysical techniques, Ground Penetrating Radar (GPR), Electrical Resistivity Tomography (ERT) and Multichannel Analysis of Surface Waves (MASW) for assessing the stability of railway earthworks. Examples of the use of each technique are provided from forensic investigations into railway earthwork slope failures in Ireland.

Significant research effort has led to improvements in our ability to estimate the ultimate bearing resistance of footings in sand. These techniques often estimate the footing resistance at relatively large displacements, typically 10% of the footing width, q _b0.1. Cone Penetration Test (CPT) design methods typically link q _b0.1 and q _c through a constant reduction factor, α. A range of α factors for shallow footings have been proposed, some methods suggest that α is constant and while others that it varies with footing width and depth (or stress level). There is a dearth of field data with which to compare these correlation factors, in particular where foundation width and depth have been varied in the same ground conditions. For this reason finite element analyses have proven to be a useful tool for performing the parametric studies required to asses factors controlling α. This paper describes the results of numerical analyses performed to investigate α factors for soil profiles which were calibrated using the results of the CPT tests performed at a dense sand test-bed site. The numerical model was first used to perform parametric analyses to consider the effect of footing width, B and footing depth, D on the α factor mobilised in dense Blessington sand. In order to assess the effects of relative density, footing tests in a range of natural sands with variable in situ densities were modeled. The results of the finite element analyses suggest that a direct correlation between q _b0.1 and q _c can be established at a given test site which is independent of footing width and depth and is relatively weakly dependent on the sands relative density if the zone of influence of the foundation considered is large enough.

The paper considers two techniques to model the Cone Penetration Test (CPT) end resistance, q_c in a dense sand deposit using commercial finite element programmes. In the first approach, Plaxis was used to perform spherical cavity expansion analyses at multiple depths. Two soil models, namely; the Mohr-Coulomb (MC) and Hardening Soil (HS) models were utilized. When calibrated using simple laboratory element tests, the HS model was found to provide good estimates of q_c. However, at shallow depths, where the over-consolidation ratio of the sand was highest, the relatively large horizontal stresses developed prevented the full development of the failure zone resulting in under-estimation of the q_c value. The second approach involved direct simulation of cone penetration using a large-strain analysis implemented in Abaqus/Explicit. The Arbitrary Lagrangian Eulerian (ALE) technique was used to prevent excessive mesh deformation. Although the Druker-Prager soil model used was not as sophisticated as the HS model, excellent agreement was achieved between the predicted and measured q_c profiles.

A geophysical investigation was carried out after the failure of an important railway embankment in the south-east of Ireland, The embankment, which had a long-term history of stability problems, was studied using a combination of ground-penetrating radar (GPR), electrical resistivity tomography (ERT), multichannel analysis of surface waves (MASW) and geotechnical testing. A significant thickening of the ballast layer around the failure location was observed using GPR, which confirmed the existence of an ongoing stability problem in the area. ERT profiles determined the presence and spatial extent of a significant layer of soft clay both beneath and to the east of the embankment, which could have a major impact on its long-term stability. ERT also detected steeply sloping bedrock close to the failure zone that is likely to have contributed to the long-term settlement of the embankment, which necessitated frequent re-ballasting. MASW confirmed the presence of the steeply sloping bedrock in addition to determining the low stiffness (G_max) values of the embankment fill. High quality sampling of the soft clay deposit was undertaken and strength and compressibility tests revealed the importance of this layer to both the on-going serviceability problems evident for the original embankment and the stability problems encountered by the remodelled section.

Research efforts have improved estimates of the ultimate base resistance of driven piles in sand. These techniques often estimate the base resistance at relatively large displacements, typically 10% of pile diameter, q _b0.1. Cone Penetration Test (CPT) design methods typically link q _b0.1 and qc through a constant reduction factor, α. Although α factors for bored piles have been proposed, some authors argue α is constant and others that it varies with footing width and depth. This paper describes the results of model tests performed using instrumented piles installed in dense sand. Numerical analyses were performed using soil profiles that were calibrated using the results of the CPT tests at the site, and used to perform parametric analyses to investigate scale effects. Results imply a direct correlation between q _b0.1 and qc, independent of pile diameter and length, suggesting model-scale testing as an ideal approach for investigating the pressure-settlement response of field-scale piles.