The MIDASclay project (Monopile Improved Design via Advanced Cyclic Soil Modelling in Clay) aims to evaluate the cyclic lateral behaviour of monopiles in clays and clay-sand mixtures, focusing on developing a cyclic soil reaction model that captures cyclic ratcheting and operational stiffness in cohesive soils. Together with the MIDAS project – aimed at cyclic behaviour of monopiles in sand, it seeks to establish a comprehensive framework for modelling the cyclic lateral response of monopiles across diverse soil conditions. To address knowledge gaps, the project employs centrifuge testing of monopiles under cyclic lateral loads, aiming to produce a soil reaction model tailored to cohesive soils. This paper outlines the project setup, research plan, and initial findings. Early results from an auxiliary centrifuge test highlight the preparation of highly over-consolidated samples and provide insights into the influence of clay properties on monopile behaviour. MIDASclay is expected to support and advance design methodologies for monopiles in cohesive soils.

Monopiles are the most common foundation for offshore wind turbines (OWTs), accounting for roughly 80% of installations in Europe. Despite advancements in research, critical knowledge gaps remain, especially regarding the behaviour of monopiles under long-term cyclic loading. Addressing these gaps is vital for enhancing the safety and costeffectiveness of future offshore wind farms and for assessing the life-cycle conditions of existing OWTs. The "MIDAS: Monopile Improved Design through Advanced Cyclic Soil Modelling" project, conducted in the Netherlands by TU Delft, Deltares, NGI, and industry partners, aimed to fill these gaps. Focusing on sandy soils, MIDAS employed a comprehensive methodology combining experimental (element and centrifuge testing), numerical, and theoretical modelling. This paper presents the centrifuge modelling component, detailing the design and implementation of the testing program, including cyclic load definitions, model pile instrumentation, loading configurations, model seabed preparation, and test procedures. Findings indicate that monotonic responses of monopiles with different configurations can be normalized effectively, validating the quality of experimental outcomes in the context of established theoretical models. The study also explores the impacts of stress levels, monopile dimensions, and drainage conditions on cyclic behaviour, contributing valuable insights to the understanding of monopile-soil interactions under cyclic loads.