Magnetic Fusion

PhD project for October 2026
Supervised by Dr Yasmin Andrew, y.andrew@ic.ac.uk 

This project is on quantitative measurements of magnetic fluctuations in the pedestal region and SOL plasmas on MAST-U and DII-D, thereby paving the way to better understanding of their roles in edge stability, transport and confinement [1].  The focus will be on cross-polarisation measurements of magnetic fluctuations between ELMs and over benign ELMs, drawing on the Scotty synthetic diagnostic.  By identifying the amplitude, wavenumber, and perhaps frequency of delta b fluctuations the underlying plasma micro instabilities can be identified and better understood.  These measurements will provide crucial insight into the electromagnetic fluctuation mechanisms [2] that regulate pedestal transport and stability.  This project will contribute to improved confinement strategies for conventional ad spherical tokamaks, potentially informing the design of future devices such as ITER and STEP.

The experimental investigation will be bolstered through interpretative numerical modelling of representative plasmas on both DIII-D and MST-u with Hermes-3 using its recent improved capability of included electromagnetic fluctuations in 3D turbulence simulations [3].

[1] AK Yeoh, VH hall-Chen, et al., 2025, arXiv: 2502.29061
[2] T Ashton-Key, Y Andrew, et al., 2025, PPCF, 67, 025027
[3] B Dudson, et al., 2024, Computer Physics Communications

Funding - TBD

PhD project for October 2026
Supervised by Dr Yasmin Andrew, y.andrew@imperial.ac.uk 

Achieving high-confinement plasma (H-mode) operation is a critical step in realising fusion conditions in next-generation devices such as ITER. The formation of the H-mode edge transport barrier (ETB) plays a central role in determining overall plasma confinement.  Recent advances in diagnostic and computational capability mean that the physics mechanisms underlying ETB formation and evolution can now be studied to a degree not previously possible.  Recent experimental studies on the DIII-D and MAST-U tokamaks, have revealed that plasma shaping, localised neutral injection, and the magnetic null-point topology exert a strong influence on edge plasma turbulence.  These effects, in turn, modify transport processes near the plasma edge and appear to be directly linked to ETB formation power thresholds.  Understanding and quantify8ing these self-regulating mechanisms is crucial to enabling reliable access to and control of the H-mode regime in future fusion devices.

This PhD project aims to develop a physics-based understanding of how magnetic shaping and neutral-plasma interactions affect edge plasma turbulence and transport, with particular emphasis on their role in ETB formation.  The project will investigate how variations in magnetic geometry, such as elongation, triangularity, and X-point position, modify edge turbulence characteristics and cross-field transport.  It will also examine the role of neutral density profile, poloidal fuelling location and localised X-point radiation in edge and scrape-off layers turbulence characteristics.  To achieve these aims the planned research will combine experimental data analysis, statistical turbulence characterisation, and gyro-kinetic simulation.  

Synthetic diagnostic outputs will be generated to allow direct comparison between simulated and experimental measurements, enabling robust validation of theoretical models.  The combined analysis will integrate theory, simulation and experiment to identify causal mechanisms by which magnetic shaping, localised radiation and neutral-plasma interaction influence turbulence and ETB formation.  Ultimately, the work will make a direct contribution to plasma confinement contorl, linking fundamental turbulence physics with fusion operation.

Funding - TBD

PhD project for October 2026
Supervised by Dr Yasmin Andrew, y.andrew@ic.ac.uk

Future burning plasma devices need to operate at high density to increase the fusion power.  The high confinement regime (H-mode) cannot always be sustained when the plasma approaches the Greenwald density limit. This is a significant challenge and limitation for high fusion performance in future devices.  Experiments show that the magnetic divertor configuration and wall material have a direct impact on the High-Density Limit (HDL), indicating that it is sensitive to the boundary plasma physics.  

This project aims to test and refine existing density limit theories using experimental data from JET and MAST-U experiments.  Existing advanced edge turbulence diagnostics on MAST-U, will be leveraged to investigate the interplay between SOL broadening, radial electric field gradients, and edge turbulence dynamics for the first time near the HDL and over H-mode to L-mode (H-L) transitions.

The project focus will be on power balance calculations, understanding the plasma self-regulation processes, and the synergy between the HDL and H-L transition.  It will also investigate the influence of impurities, and the X-point radiator regime to confront and develop physics understanding of the underlying processes. Turbulence simulations are currently being carried out using Hermes-3 on the ARCHER supercomputer and will be used in this project to complement experimental analysis and develop synthetic diagnostic data.  The outcomes will inform strategies to extend H-mode operation to higher densitites in future reactors and is based on current research at Imperial and UKAEA.

Funding - TBD

PhD project for October 2026
Supervised by Dr Robert Kingham, rj.kingham@ic.ac.uk 

Compact spherical tokamaks like MAST-U have limited space to reduce exhaust fluxes, and so use a combination of methods such as advanced divertor configurations and detachment via impurity seeding.  The current leading design is the connected double null (CDN) magnetic configuration which is up-down symmetric  The design aim is for the upper- and lower-outer divertors to receive most of the parallel heat flux, while the smaller inner ones receive less.  Recent fluid simulations (SOLPS-EIRENE) show that even in well-balanced CDN plasmas, there are asymmetries (top-down and inner-outer) in power loading [1] despite the absence of drifts known to driver asymmetry [2].  In future higher-power devices and reactors, kinetic effects are expected to modify collisional transport beyond the capabilities of fluid codes used for exhaust design, with implications on the real performance of predicted designs.

This project aims to improve our current kinetic modelling of parallel transport with ReMKiT1D [3] - a framework for creating kinetic and fluid exhaust models of parallel transport - to study this asymmetry  It will leverage the flux-tube expansion capability being added by a current Imperial PhD student.

Addition of kinetic ions: Currently, ReMKiT1D only treats electrons kinetically, with ion species treated as fluids with classical closures.  It is proposed to extend the kinetic framework to the hydrogenic ion species (D and T) as well. A key challenge will be tractable treatment of the ion-ion Coulomb collision operator in the spherical-harmonic velocity space approach used in ReMKiT1D.  A suitable simplified form will be sought and its characteristics assessed.  This capability would circumvent the need for a collisional multi-species ion closure (e.g. Zhdanov or Makarov [4]), allow the development of non-Maxwellian ion distribution functions during acceleration in the pre-sheath, allow the D and T ions to differentially diffuse and permit high-energy ions to stream from upstream to the targets under the lower collisionality conditions anticipated in reactor-scale devices.  The influence of such effects on collisional multi-species closures, sheath transmission coefficients and partitioning amongst energy channels at the inner and outer divertors are interesting and important gaps in understanding that can be investigated.

[1] Osawa RT, et al. Nucl. Fusion 63, 076032 (2025)
[2] Paradela Pérez I, et al., Nucl. Fusion 65, 066026 (2025)
[3] Mijin S, et al., Comp. Phys. Comm. 300, 109195 (2024)
[4] Makarov SO, et al., Phys. Plasmas 28, 062308 (2021)

Funding - TBD