Physics-Informed Surrogate Modeling for Offshore Helical Pile Design Optimization

The development of offshore wind turbine (OWT) foundations toward larger sizes and deeper embedment presents challenges for traditional foundations, which have long installation times and cause significant environmental disturbances.

Helical piles, with their distinct installation method, offer shorter installation times and reduced environmental impact, making them a strong candidate for OWT foundations.

However, the lack of efficient methods to assess offshore HP load-bearing performance under various conditions limits their broader application.This project aims to develop an open-source decision-making tool to optimize offshore HP design using data-driven techniques.

By incorporating physics-informed machine learning algorithms (PIA), the tool will enable surrogate modeling to replace traditional numerical analyses, allowing accurate predictions of HP’s static and dynamic performance, including installation effects.

This will improve design processes, reduce costs, and contribute to sustainable offshore engineering practices.The fellowship has three key Research and Innovation Objectives (R&IO): (1) to establish a validated numerical model that simulates HP load-bearing performance, including installation effects, which will form a database for (2) developing PIA-based data-driven models to assess HP capacity.

These models will then be integrated into (3) an open-source decision-making tool using reliability-based foundation design approach (RBFDA) to optimize HP design under various conditions.The fellowship will be conducted at NGI, Norway, with a secondment at OsloMet, Norway.

The skills gained will enable the researcher to become an expert in HP technology and support the pursuit of a tenure-track position.

Furthermore, the project aligns with the European Green Deal and the UN Sustainable Development Goals (SDGs) by contributing to clean energy and sustainable engineering.