Machine Learning-based Optimisation of Relativistic Laser–Plasma Coupling for Efficient Fusion Ignition

About the Project

Fusion research has recently reached major milestones, showing that laser‑driven fusion could one day become a practical clean‑energy source. To make further progress, we need to understand and control how powerful lasers transfer energy to plasma. This PhD focuses on developing new ways to predict and optimise this laser–plasma coupling using experiments, advanced simulations, and modern machine‑learning techniques.

What the project is about

This project will investigate how high‑power laser pulses interact with plasma at relativistic intensities. These interactions determine the efficiency of energy transfer and shape the characteristics of the resulting beam of high‑energy electrons—critical factors for advanced ignition approaches to fusion.

The project brings together several exciting areas:

• High‑power laser experiments at world‑class facilities (EPAC and SCAPA) where you will measure how different laser conditions affect energy absorption and particle acceleration.
• Particle‑in‑cell (PIC) simulations using the EPOCH code to model these interactions in detail.
• Machine learning and Bayesian optimisation to build predictive models that can identify the best laser and target parameters without the need for large numbers of costly simulations or experiments.
• Transfer learning approaches that link short‑pulse experiments to longer‑pulse fusion‑relevant conditions.

What you will do

• Explore how laser‑pulse properties—such as polarisation, angle of incidence, and temporal‑intensity contrast—as well as pre‑plasma conditions influence energy absorption and electron acceleration.
• Combine simulations with experimental data to train surrogate models capable of rapidly predicting key outcomes.
• Develop an optimisation framework to guide future fusion experiments and support improvements to advanced ignition strategies.

Why this project matters

Better control of laser‑plasma coupling is essential for developing reliable, high‑yield fusion interactions. The predictive framework created in this project will support the design of next‑generation experiments on facilities such as the new Vulcan 20‑20 laser, helping accelerate UK and international efforts towards practical fusion energy.

Training and environment

You will join a vibrant laser–plasma physics group at Strathclyde and collaborate with leading researchers at the Central Laser Facility at the Rutherford Appleton Laboratory. You will also have access to:

• State-of-the-art high power laser facilities, including the SCAPA facility at Strathclyde
• National high‑performance computing facilities
• Training in machine learning, simulation, and plasma physics through SUPA and the University’s PGCert‑Researcher Professional Development programme

Your supervisory team includes leading experts in experimental laser–plasma physics, computational modelling, and machine‑learning‑driven optimisation.

Who should apply

PhD candidates must hold at least an upper Second‑Class UK Honours degree, or an international equivalent, in physics. Experience in plasma physics and/or machine learning is desirable, while enthusiasm for learning and a strong interest in combining physical science with advanced computation are essential.

Potential applicants are encouraged to contact Prof. Paul McKenna (paul.mckenna@strath.ac.uk) for more information.

How to apply

Apply on line via the Strathclyde Physics Application Portal or directly here: https://pegasus.mis.strath.ac.uk/applicantentry/#/?ppio=287208