PhD Opportunity: Machine Learning for CFD of Point Particles (FlowLB)

We invite self-funded applicants to pursue a PhD at Newcastle University integrating machine learning with computational fluid dynamics for particle-laden flows using the lattice Boltzmann solver FlowLB. The project develops data-driven closures for point- particle models, trained on high-fidelity particle-resolved simulations and LES datasets, to improve predictions of momentum and heat transfer while reducing computational cost.

Research themes include training ML models for drag, lift, and heat transfer; embedding surrogate models into LBM simulations for online coupling; and assessing model generalisability across Reynolds numbers, particle sizes, and flow regimes. Candidates will use HPC resources and collaborate with international partners, contributing to publications and conference presentations. The successful candidate will gain hands-on experience in high- performance computing, model validation, uncertainty quantification, and reproducible research practices, enhancing prospects.

Applicants should have a strong background in CFD, numerical methods, and programming. Experience with machine learning frameworks is advantageous. This project offers a unique opportunity to advance hybrid ML–physics modelling for complex multiphase flows and to develop transferable tools for engineering and environmental applications.