PHD Position Machine-learning Methods for Scale

PhD Position Machine-Learning Methods for Scale-Resolving Simulations of Turbulent Flows Delft University of Technology (TU Delft) Delft, Netherlands EUR 3,059-3,881 / month Posted on Nov 11, 2025 Job description The exponential increase of computational power over the last decade has enabled scale-resolving simulations (SRS) of turbulent flows at an unprecedented resolution. In combination with high-performance computing (HPC), parallel computational fluid dynamic simulations can currently resolve spatial and temporal scales of industrially-relevant turbulent flows within days/weeks. On the other hand, typical design cycles in industry still make extensive use of low-fidelity simulations which can provide fast but inaccurate solutions depending on the flow complexity. To close this gap, this PhD will explore machine-learning (ML) methods to significantly reduce the turnaround time of SRS, thus enabling their use for industrial design processes. By combining state-of-the-art numerical

methods and data-driven modelling techniques, the PhD candidate will investigate novel frameworks to accelerate SRS in HPC environments. The PhD candidate will work at the Ship Hydrodynamics section of the Department of Maritime and Transport Technology (Faculty of Mechanical Engineering). This environment facilitates the exchange of research knowledge with colleagues also working in CFD and ML, both in applied and fundamental areas. The candidate will develop transferable skills such as HPC software development, extensive physical understanding of turbulent flows, and experience in modern ML architectures such as neural operators and transformers. The expected start date is February 2026, and the PhD duration contract is 4 years. Please contact Dr. Bernat Font ( b.font@tudelft.nl ) for more information. Job requirements - MSc in Mechanical/Maritime/Aeronautical Engineering, Physics, Applied Mathematics, or related discipline. - Proven track r...