ÌìÃÀ´«Ã½

We have multiple research projects ongoing in Offshore Geotechnics, covering the long-term response of soils and foundations under repeated loading, seismic performance of offshore foundations, pile installation behaviour, and the application of data-driven methods to offshore foundation design. Much of this work is motivated by the rapid expansion of offshore wind energy and the need for more reliable, computationally efficient design tools for monopile and other foundation systems in complex ground conditions.

Modelling the Response of Laterally Loaded Piles

A central strand of our offshore research concerns the prediction of long-term deformation accumulation in pile foundations subjected to the sustained cyclic lateral loading imposed by waves and wind. Conventional cycle-by-cycle finite element analyses are prohibitively expensive for the tens of thousands of cycles relevant to offshore wind turbine design, and the High-Cycle Accumulation (HCA) framework offers a practical alternative. We validated an implementation of HCA coupled with a practice-oriented sand constitutive model in PLAXIS, described in detail in  in MethodsX. A rigorous calibration procedure using only site-specific investigation and laboratory data was established, and the approach was validated against the monotonic and cyclic lateral field tests from the PISA Joint Industry Project at Dunkirk, France, including lateral pile load tests up to 30,000 cycles, in  in Géotechnique. Earlier work also examined the sensitivity of 3D finite element monopile pushover analyses to natural variability in ground characterisation (), with recent work providing further insights (Tantivangphaisal, Taborda and Kontoe, 2026) when using HCA models in practice and a downloadable PLAXIS user defined soil model provided in .

Recent conference contributions have extended this work to the evaluation of multi-directional and combined force-moment limit state envelopes for monopile design in sands (), and to the GEOLAB international blind prediction contest for laterally loaded piles under monotonic and cyclic loading, in which our team achieved the best overall performance across both loading scenarios (). The effect of small-strain stiffness on the lateral behaviour of monopile foundations has also been investigated in collaboration with colleagues from the University of Thessaloniki ().

Seismic Performance and Liquefaction

The seismic response of offshore foundations presents distinct challenges compared to onshore structures, including assessment of liquefaction at depth and the significance of vertical ground motion in offshore environments.  examined these challenges through 3D dynamic finite element analyses of a 5 MW turbine on a monopile foundation, demonstrating that resonant frequencies can be excited under realistic ground motions and that the water column must be included for accurate representation of the soil-water system's compression natural frequency. The combination of kinematic and inertial loads on monopile foundations was further explored by , and a broader overview of dynamic analysis across pile installation and seismic performance was presented as a keynote by  at SECED 2023.

The constitutive modelling of sand under repeated seismic loading — including post-liquefaction response and the effects of prior shaking history — has been assessed by , while the response of shallow foundations on liquefiable deposits, with application to Wind Turbine Installation Vessels, was investigated by .

A closely related strand of work addresses the modelling of soil nonlinearity in seismic soil-structure interaction analyses. compared two approaches for assessing the seismic response of a 15 MW offshore reference wind turbine on a 10 m diameter, 45 m embedded monopile in clay: a cyclic nonlinear constitutive model with enhanced hysteresis control (IC MAGE M04, implemented as a user-defined model in PLAXIS 3D) and a simplified equivalent linear (EQL) visco-elastic model informed by 1D site response analyses. The EQL approach consistently predicted higher accelerations and bending moments than the nonlinear model, with discrepancies in maximum bending moment of around 10% for a standard-intensity Kobe earthquake record and up to 23% for a downscaled version. A key finding concerns resonance: when the natural frequency of the soil deposit aligns with the tower's second mode, EQL models are particularly susceptible to sustained amplification because their soil properties remain constant throughout the motion, whereas the nonlinear model's continuously evolving stiffness prevents the system from being locked into a single resonant frequency. The study concludes that EQL analysis is a suitable conservative first approximation for design, but that discrepancies grow with seismic intensity and that further investigation is needed for different soil types and ground motion characteristics.

Pile Installation in Chalk

Chalk is encountered widely at potential offshore wind development sites across Northern Europe, and impact driving in low-to-medium density chalk poses distinctive challenges: it de-structures the material close to the pile shaft to form a thin annulus of 'putty' chalk, induces additional fracturing in the surrounding rock, and generates excess pore pressures that govern the time-dependent build-up of axial capacity during ageing. Our research contributes to this area through a sustained programme of numerical modelling alongside the ALPACA and ALPACA Plus Joint Industry Projects.

published finite element analyses in Computers and Geotechnics of open steel tubular piles driven in low-to-medium density chalk under monotonic axial loading. The study characterised the role of the de-structured chalk annulus, the influence of ageing on radial effective stresses, and the potential for strain-softening in this brittle material. The analyses showed that accounting for the ALPACA-SNW design guidance-derived radial stress profiles is essential to reproduce observed capacity, and that compression shaft capacity substantially exceeds tension capacity.

Building on this axial work, presented coupled hydro-mechanical finite element simulations of impact hammering on open-ended piles in chalk, published in Soil Dynamics and Earthquake Engineering. Using measured force-time histories from dynamic sensors near the pile head as direct input, and adopting a zoning approach to capture the radially varying chalk properties, the analyses demonstrate that the puttified chalk zone governs dynamic pile response, and that the framework can capture both the excess pore pressures generated during continuous percussive driving and the time-dependent axial resistances mobilised in restrike hammer blows.

Lateral loading behaviour has been addressed in a companion journal paper, , which extends earlier 3D finite element analyses to a wider range of pile geometries tested under the ALPACA and ALPACA Plus programmes — including piles up to 1.22 m in diameter. The study shows that the simplified approach of assigning degraded mechanical properties to the de-structured and fractured zones around the pile, originally validated on smaller piles, transfers well to larger geometries when pile steel yielding is explicitly accounted for. However, predictions of early-stage lateral stiffness remain sensitive to assumptions about the extent and stiffness of the damaged Zone B, highlighting the need for improved chalk characterisation in this region. The conference counterpart of this work was also presented at ISFOG2025 ().

Offshore Foundations and Data-Driven Design

We have also investigated the behaviour of suction bucket foundations under axial loading through three-dimensional numerical modelling (), and through machine learning.  presented a data-driven macroelement model for suction buckets in sand at ISFOG2025, offering a computationally efficient alternative to full three-dimensional finite element analyses for use in design workflows. Data-driven approaches are more broadly applied across offshore foundation problems:  explored the use of machine learning for back-analysing monopile response,  applied data assimilation techniques to the serviceability assessment of offshore wind foundations, and Yang et al. (2024) developed a surrogate model for monopile design using machine learning. These contributions reflect the group's broader commitment to integrating advanced computational and data-centric methods into offshore geotechnical practice.

Full set of publications and open datasets are available through our page and the .