Effective stress analysis of residual wave-induced liquefaction around caisson-foundations: Bearing capacity degradation and an AI-based framework for predicting settlement

Abstract

Excessive wave-induced pore pressure buildup around offshore foundations results in liquefaction, settlement and bearing capacity degradation, which may threaten the safety of Offshore Wind Turbines (OWTs). Despite the extensive research efforts on wave-induced seabed residual response in the recent years, there is still a lack of knowledge on mechanisms of wave-induced liquefaction and settlement around caisson foundations. Employing a code-based framework implemented in OpenSees, the cyclic response of wave-seabed-foundation (WSF) system is evaluated. Biot’s consolidation theory, linear wave theory and CycLiqCPSP constitutive soil model are integrated to evaluate the response of soil-foundation system accounting for the hydrodynamic pressure of wave imposed on the seabed surface considering the fully-coupled wave-seabed-foundation interaction (WSFI). Soil model parameters are calibrated against cyclic simple shear (CSS) tests and the numerical model is validated by a well-documented centrifuge experimental model. Various seabed characteristics, and a range of the geometrical properties of foundation with different OWT weights are put into practice for evaluating the geotechnical aspects of wave-induced liquefaction. Stress paths are provided to demonstrate quite different level of reduction in effective stresses and likelihood of liquefaction occurrence upon cyclic shear stresses application for different locations in the vicinity of the foundation area. The present study will also improve understanding of the interplay among the state variables on the wave-induced foundation settlement and bearing capacity degradation allow to feed the output data into development of simplified procedure for assessing the bearing capacity. Finally, results from over 250 analyses with different model configurations are used to provide an estimate of wave-induced caisson settlement with reasonable accuracy on the basis of artificial intelligence (AI) method known as Group Method of Data Handling (GMDH).