Research
We investigate how fuelling, feedback and dynamic plasma behaviour can be treated as active control variables in fusion systems. Our work spans phase-referenced fuel delivery, strongly self-heated plasma dynamics and simulation-led control research, using progressively higher-fidelity models to test behaviour, expose failure modes and build reproducible evidence.
WAFI
Wake-Aligned Fuel Injection (WAFI) investigates whether fusion-fuel deposition referenced to the phase of a propagating fusion-reaction structure can influence system behaviour. The research focuses on estimating structure phase, selecting a target deposition relationship and controlling fuel delivery relative to that evolving state.
Reduced-order studies have shown phase-dependent behaviour, while higher-dimensional reduced-MHD work has produced provisional results consistent with that signal. Current work is focused on understanding the limits, robustness and physical conditions under which the effect persists rather than assuming a particular reactor architecture.
High-Q Plasma Dynamics
We are developing a high-fidelity 3D research environment to investigate local plasma behaviour under strongly self-heated fusion conditions. By deliberately restricting the simulated physical domain, computational resources can be concentrated on spatial and temporal fidelity rather than whole-device scale.
The initial question is deliberately open: what structures, transport behaviour and local responses emerge naturally under these conditions? No WAFI control mechanism or predetermined propagating structure is imposed. Interesting behaviour can then be characterised before any question of control is introduced.
Simulation-led, evidence-driven
Research progresses from reduced-order models into multidimensional plasma simulation, control-oriented diagnostics and progressively higher-fidelity testing. Live telemetry and dedicated analysis tooling are used to inspect behaviour during active runs, while sensitivity, convergence and validation testing are used to distinguish persistent physical behaviour from numerical artefact.