PhD Position Neural Network-Based Surrogate Models for Offshore Wind Foundations

Join the RAPID-Wind project to develop reduced-order and surrogate models with neural operators and geometric deep learning for offshore wind foundations

Job description

We are looking for a highly motivated and talented PhD researcher to join the RAPID-Wind project, which aims to develop a new computational framework for the design of advanced offshore wind turbine foundations in deep waters. As turbine sizes increase and installations move to greater depths, the offshore industry faces growing challenges related to wave loading, dynamic response, and fatigue. Perforated monopiles are a promising concept to reduce hydrodynamic loads and increase passive damping, but their design requires fast and reliable prediction tools that can approximate complex free-surface, multiscale flow–structure interactions at a fraction of the cost of high-fidelity simulations. RAPID-Wind will develop a new computational modelling framework that enables high-fidelity simulations and near real-time predictions by combining adaptive numerical methods, high-performance computing (HPC), and efficient surrogate models based on reduced-order modelling (ROM) and neural operators.

Your role

If selected, you will focus on developing reduced-order and surrogate models for the fast and accurate prediction of hydrodynamic loads and stress distributions in perforated offshore structures. The research will emphasize data-driven and learning-based ROM approaches for complex free-surface flow problems, combining numerical modeling with modern machine learning techniques. You will work with surrogate modeling concepts such as neural networks and neural operators, including approaches from geometric deep learning for handling complex geometries. In addition, you will investigate multi-fidelity and physics-informed training strategies to ensure robust and reliable predictions in data-scarce regimes. The reduced-order models developed in this PhD project will form a central building block for fast prediction and design exploration within the overall RAPID-Wind framework.

Research environment

You will join the Numerical Analysis section at the Delft Institute of Applied Mathematics, in particular the SCaLA (Scalable Scientific Computing and Learning Algorithms) group, which develops scalable numerical methods for partial differential equations, reduced-order modeling, and scientific machine learning, with a strong focus on complex geometries and high-performance computing on modern CPU and GPU architectures. Your PhD project will be co-supervised by Alexander Heinlein and Oriol Colomés, lead of the Computational Multiphysics in Offshore Engineering (CMOE) group. You will work in close collaboration with the Offshore Engineering section in the Hydraulic Engineering Department, actively participate in regular group meetings, publish scientific articles, present your work at national and international conferences, and contribute to teaching and supervision activities within the Faculty of Electrical Engineering, Mathematics and Computer Science at Delft University of Technology.

A key aspect of this PhD project is close collaboration with industry partners to ensure that the research translates into real-world design practice. The research will be conducted in cooperation with companies and organizations leading the design and analysis of offshore wind foundations, and the definition of datasets and output quantities of interest for the reduced-order and surrogate models will be carried out jointly with these partners.

Job requirements

The successful applicant will have:

• A master’s degree in Applied Mathematics, Data Science, Machine Learning, Computer Science, Engineering, or another closely related field.
• Demonstrated expertise in machine learning and scientific computing, with a solid background in numerical methods for differential equations; experience with neural networks, surrogate modeling, reduced-order modeling, or scientific machine learning is expected.
• Strong programming skills in Python, with experience using modern machine learning frameworks such as PyTorch or JAX; experience with C++ or Julia is an advantage.
• A self-motivated, curiosity-driven mindset and openness to communication and collaboration.
• Excellent communication skills in English, both written and spoken.
• The ability to work independently and proactively within a multidisciplinary research team.

About TU Delft as an employer

TU Delft (Delft University of Technology) is a top international university combining science, engineering and design.

Conditions of employment

Doctoral candidates will be offered a 4-year period of employment in principle, but in the form of 2 employment contracts. An initial 1.5 year contract with an official go/no go progress assessment within 15 months. Followed by an additional contract for the remaining 2.5 years assuming everything goes well and performance requirements are met.

As a PhD candidate you will be enrolled in the TU Delft Graduate School. The TU Delft Graduate School provides an inspiring research environment with an excellent team of supervisors, academic staff and a mentor. The Doctoral Education Programme is aimed at developing your transferable, discipline-related and research skills.

The TU Delft offers a customisable compensation package, discounts on health insurance, and a monthly work costs contribution. Flexible work schedules can be arranged.