Oxford Joins Imperial and York in £12 Million Fusion Power Industry Partnership

A multi-institutional team including Trinity Physics Fellow Sam Vinko will explore a new method for creating fusion power it hopes could eventually be scaled to provide safe, clean, and abundant energy. Support for the project comes from a combined £12 million from First Light Fusion, the company behind this new approach, Machine Discovery Ltd., an Oxford AI startup, and UK Research and Innovation’s Prosperity Partnership scheme.

The project will see Oxford researchers from the Departments of Engineering Science and Physics join forces with colleagues from Imperial College London, University of York, and the industrial partners to investigate specific phenomena relating to hydrodynamics and heat transport using X-ray imaging techniques they have developed over the years.

Nuclear fusion occurs when the nuclei of two atoms, for example hydrogen atoms, are combined to create a different element such as helium, releasing a huge amount of spare energy due to the difference in weight between the atomic ingredients and the newly created atom. Fusion is known to have transformational potential as a safe, clean, and abundant energy source if fusion conditions, which require intense heat and pressure, could be created. However, techniques tried so far have only generated limited amounts of energy, leading to questions about the scalability of such methods and of fusion power in general.

Under the new partnership, researchers from across the three universities and two companies will work together to study the behaviour of materials at extreme temperatures, pressures and densities, examining how heat, matter, and radiation flow at interfaces between those materials.

Dr Vinko’s role in this collaborative effort will be to perform experiments exploring the behaviour of the extreme states of matter needed for inertial fusion energy science. Much of this work will be performed at large-scale laser and free-electron laser facilities worldwide, and will be combined with both computational physics and novel developments in machine learning.

‘Machine learning tools, such as neural-network-based emulators, have a particularly important role to play in exploring the extreme states of matter required to deliver fusion,’ he explains, ’as interpreting complex experiments is challenging, and the modelling requirements often exceed our computational capabilities. This partnership brings together the wide range of expertise that is needed to start addressing this grand challenge.'