A modification of the high-cycle accumulation (HCA) framework coupled with a practical constitutive model for sands and its numerical implementation as a user-defined soil model in PLAXIS is presented. The implemented model is compared against data from the original high-cyclic tests in Karlsruhe fine sand and more recent laboratory tests in Dunkirk sand. A reference 15 MW offshore wind turbine monopile foundation subject to lateral cyclic wave loading is used in an engineering design scenario at three different load levels to verify the current numerical implementation. Details include: center dot Modifications made to the HCA framework to couple it with a practical sand constitutive model, center dot Implementation of an efficient workflow to switch between low and high cycle constitutive equations in PLAXIS, and center dot Verification of the implementation at single element and boundary value problem scales.

The offshore wind turbines (OWT) are subjected to cyclic loads, such as ocean waves and wind, over extended periods. The soil surrounding the pile experiences bi-directional cyclic shear. As a result of the low-frequency and long-term loading in the pile-soil interaction, the cumulative deformation of pile foundation increases, posing a risk to the operational safety of wind turbine system. The soil around the piles is distributed with soft clay and clay layers. To study the cumulative deformation properties of clay under complex stress states. A series of tests are conducted, the variation of resilient modulus under different cyclic stress levels and confining pressures is analyzed based on test results. Then an empirical model uniformly reflecting strain-hardening and strainsoftening properties of clay is proposed. The variations of model parameters are investigated. Then the established empirical model is used to modify the maximum elastoplastic modulus at each unloading within the bounding surface constitutive model, a parameter reflecting the magnitude and rate of strain accumulation is also introduced. This method is characterized by a simple expression and requires fewer model parameters. Finally, the predicted results of modified constitutive model are compared with test results to verify the validity of the established model.