PhD Studentship: Mechanical Engineering, Fusion, Digital: AI Enhanced Modelling of Tokamak-type Fusion Power

A fusion power plant is an exceptionally complex multidisciplinary system comprising highly coupled components and interacting physical processes, including magnets, breeder blankets, divertors, plasma systems, and structural materials. The underlying physics spans electromagnetics, thermo-mechanics, plasma dynamics, neutronics, material degradation, and heat transfer. In addition, the manufacturing, construction, and operational challenges associated with tokamak-based systems are substantial.

For sustainable fusion power generation, these tightly coupled systems must operate seamlessly, reliably, and predictably under extreme conditions. This represents not only a major design challenge but also a large-scale optimisation problem due to the strong interdependence between components, materials, and physical processes. At the current stage of fusion development, there is a pressing need for integrated whole-plant modelling frameworks capable of incorporating emerging experimental data, operational knowledge, and multiphysics interactions.

This project will undertake a comprehensive systems-level analysis, starting with two interdependent disciplines, before adding additional disciplines. A hierarchy of computational approaches, including lumped-parameter, reduced-order, and high-fidelity multiphysics models, will be explored. Particular emphasis will be placed on uncovering the relationship between high- and low-fidelity models to uncover, for example, loss of accuracy, speed and/or robustness. Additionally, optimisation and uncertainty quantification techniques with the potential to be applied across the fidelity spectrum will be investigated.

The project will also integrate state-of-the-art machine learning and AI methodologies to accelerate design optimisation, uncertainty quantification, and data integration. Ultimately, the research aims to establish a next-generation federated digital platform capable of linking multiple disciplines and data streams into a unified predictive framework for future fusion power plants.

Applications may be submitted in Welsh and any application submitted in Welsh will be treated no less favourably than an application submitted in English. Please refer to the University's Welsh Language Policy on Awarding Grants.

Funding Details

Covers full tuition, £21,805 stipend (2026/27), plus up to £1,000 yearly for research costs.