PhD Project: Machine Learning Investigation of Unresolved Point Particles in Turbulent Flows (self-funded applicants)

The accurate representation of unresolved point particles remains a fundamental challenge in multiphase flow simulations. Traditional models often rely on empirical assumptions that limit predictive capabilities, especially in complex or turbulent environments. This PhD project aims to revolutionize particle-fluid interaction modeling by leveraging advanced machine learning techniques in CFD. By analyzing high-fidelity simulation data and experimental datasets, the project will develop data-driven models that can accurately predict particle forces, trajectories, and interaction dynamics within the fluid without relying solely on simplified closures.

The candidate will explore a range of machine learning approaches, including neural networks, ensemble methods, and physics-informed models, to capture hidden patterns in unresolved particle behavior. Emphasis will be placed on ensuring physical consistency, generalization across different flow regimes, and interpretability of the resulting models. Applications include environmental flows, industrial particulate processes, and biomedical systems where particulate transport is critical.

This project is based at Newcastle University, a prestigious institution ranked 129th in the world according to the QS World University Ranking. Newcastle offers access to excellent research facilities, a vibrant academic community, and a supportive environment for doctoral researchers.

Applicant Requirements

• Strong motivation for interdisciplinary research.
• Excellent analytical and problem-solving skills.
• Background in computational fluid dynamics (CFD) or related fields.
• Preferred: Experience with machine learning (ML) techniques.
• Preferred: Familiarity with programming languages such as Python, MATLAB, or C++.
• Good communication skills and ability to work both independently and as part of a team.