Robert S. Granetz

The development of robust machine learning models for accurate plasma transient analysis and prediction necessitates large-scale, consistent datasets of equilibrium reconstructions with high temporal resolution. Traditional post-discharge workflows for equilibrium reconstruction such as EFIT are typically device-specific or limited to relatively coarse time grids, creating significant bottlenecks for data-driven fusion research. To address this, we present DisruptionEFIT, an automated, tokamak-agnostic framework designed to perform large-scale EFIT recomputation at 1 kHz sampling rate. Built on a modular Python architecture, the system orchestrates the selection, execution, and validation of equilibrium reconstructions across diverse hardware environments, supporting both the Alcator C-Mod and DIII-D tokamaks. The framework integrates seamlessly with MDSplus for data ingestion and utilizes device-specific backends to manage parallel execution via multiprocessing or SLURM batch scheduling. Key features include automated error recovery, rigorous quality assurance through real-time worker validation and post-hoc verification checks, and the injection of reproducible metadata into output trees. We present results from successful deployment on both tokamaks, demonstrating efficient processing of large-scale datasets with robust quality control and full traceability. By standardizing the reconstruction process and ensuring time-uniform data grids, DisruptionEFIT facilitates the generation of high-fidelity datasets essential for advancing disruption prevention strategies and AI-driven plasma control solutions.

This Roadmap article addresses the critical and multifaceted challenge of plasma-facing component (PFC) damage caused by runaway electrons (REs) in tokamaks, a phenomenon that poses a significant threat to the viability and longevity of future fusion reactors such as ITER and DEMO. The dramatically increased RE production expected in future high-current tokamaks makes it very difficult to avoid or mitigate REs in such devices when a plasma discharge terminates abnormally. Preventing damage from the intense localized heat loads they can cause requires a holistic approach that considers plasma, REs and PFC damage. Despite decades of progress in understanding the physics of REs and the thermomechanical response of PFCs separately, their complex interplay remains poorly understood. This document aims to initiate a coordinated, interdisciplinary approach to bridge this gap by reviewing experimental evidence, advancing diagnostic capabilities, and improving modeling tools across different scales, dimensionalities, and fidelities. Key topics include RE beam formation and transport, damage mechanisms in both brittle and metallic PFCs, and observed effects in major facilities such as JET, DIII-D, WEST and EAST. The Roadmap emphasizes the urgency of predictive, high-fidelity modeling validated against well-diagnosed controlled experiments, particularly in the light of recent changes in ITER’s wall material strategy and the growing importance of private sector fusion initiatives. The modeling gaps include calculations of the magnetic equilibrium, RE distribution function and background plasma properties at the start of the beam termination, with a critical need to advance modeling of access conditions to benign terminations. This is to be followed by simulations of the termination itself which requires a self-consistent description of the 3D MHD fields together with the REs, which is now available thanks to recent developments. Finally, the damage assessment is enabled by modeling of the full thermo-mechanical response, encompassing the hydrodynamic and the deviatoric behavior, to RE impacts with characteristics as provided by the termination simulations. The full workflow is computationally heavy for scoping studies to assist design of future machines, calling for development of reduced or surrogate models. Each section of the Roadmap article is written to provide a concise overview of one area of this multidisciplinary subject, with an assessment of the status, a look at current and future challenges, and a brief summary. The ultimate goal of this initiative is to guide future mitigation strategies and design resilient components that can withstand the intense localized loads imposed by REs, thus ensuring the safe and sustainable operation of the next generation of fusion power plants.

Plasma turbulence measurements play a crucial role in understanding transport, which affects the fusion yield of tokamaks. However, there is a notable scarcity of plasma turbulence diagnostics, particularly temperature fluctuations. While soft x-ray (SXR) imaging holds promise for measuring electron temperature fluctuations, such a fluctuation diagnostic has not yet been developed. In our prior work (Chen et al 2021 Rev. Sci. Instrum. 92 055337), we successfully designed an SXR-based electron temperature fluctuations diagnostic and developed a reconstruction algorithm called pseudolocal tomography. In this paper, we further demonstrate the feasibility of this diagnostic using fluctuation data directly derived from gyrokinetic simulations of NSTX-U L-mode discharges. The reconstruction algorithm has been refined to achieve the same accuracy using a simpler diagnostic layout than in previous work (Chen et al 2021 Rev. Sci. Instrum. 92 055337). The enhanced algorithm also enables the reconstruction of electron density fluctuations. In light of this, we prefer to refer to this tool as a ‘fluctuation diagnostic’ rather than solely an ‘electron temperature fluctuation diagnostic’. Consistent with our previous work, this study explores the diagnostic’s range of detection, emphasizing its ability to reconstruct the 2D k – f spectra of both electron density and temperature fluctuations in both radial and poloidal directions, and the cross angle phase between them. Subsequently, we investigate how the diagnostic’s performance is influenced by the region of interest (ROI), diagnostic parameters such as the number of lines of sight and viewing angles, signal-to-noise ratio and finite beam size. Through comprehensive analysis, we establish the feasibility of this new SXR diagnostic to measure turbulent fluctuations, including relevant potential diagnostic limitations such as the diminished effectiveness when the ROI is too close to the plasma edge.

This Roadmap article addresses the critical and multifaceted challenge of plasma-facing component (PFC) damage caused by runaway electrons (REs) in tokamaks, a phenomenon that poses a significant threat to the viability and longevity of future fusion reactors such as ITER and DEMO. The dramatically increased RE production expected in future high-current tokamaks makes it difficult to avoid or mitigate REs when a plasma discharge terminates abnormally. Preventing damage from the intense localised heat loads REs can cause requires a holistic approach that considers plasma, REs and PFC damage. Despite decades of progress in understanding the physics of REs and the thermomechanical response of PFCs, their complex interplay remains poorly understood. This document aims to initiate a coordinated, interdisciplinary approach to bridge this gap by reviewing experimental evidence, advancing diagnostic capabilities, and improving modelling tools across different scales, dimensionalities and fidelities. Key topics include RE beam formation and transport, damage mechanisms in brittle and metallic PFCs, and observations in major facilities such as JET, DIII-D, WEST and EAST. The Roadmap emphasises the urgency of predictive, high-fidelity modelling validated against well-diagnosed controlled experiments, particularly in the light of recent changes in ITER's wall material strategy and the growing importance of private sector initiatives. Each section of the article is written to provide a concise overview of one area of this multidisciplinary subject, with an assessment of the status, a look at current and future challenges, and a brief summary. The ultimate goal of this initiative is to guide future mitigation strategies and design resilient components that can withstand the loads imposed by REs, thus ensuring the safe and sustainable operation of the next generation of fusion power plants.

Previous investigations on JET suggest half or less of plasma stored thermal energy Wth is radiated (frad,th≲0.5) using either massive gas injection (MGI) or shattered pellet injection (SPI) disruption mitigation. We investigate whether the apparent incomplete frad,th is explained by radiation peaking near the injection plume. High toroidal peaking throughout the pre-thermal quench is found in argon–deuterium MGI on JET, with typically >3× higher radiation near the injector than toroidally distant. Previously unexplained toroidal bolometry measurements in neon–deuterium SPI are reproduced with similar peaking using the Emis3D radiation analysis code. These observations align with results from Alcator C-Mod and KSTAR. This peaking is not captured by previous JET studies that found poor thermal mitigation. Two sets of neon–deuterium SPI and two sets of argon–deuterium MGI are analyzed using Emis3D. In SPI, frad,th rises from no-plume estimates of 0.31 and 0.66 to lower bounds of 0.84 and 0.92, respectively, and frad,th∼1 is possible. In MGI, the toroidal spread of the peaking feature is poorly constrained. frad,th up to 0.85 and 0.65 are possible using the largest possible spread, increasing from 0.42 and 0.28, although frad,th∼1 does not appear to be reached. Revised mitigation estimates on JET suggest a lower melt risk to the divertor in mitigated disruptions on ITER and SPARC than previously thought. However, peaking near injectors could increase flash melting risk on nearby plasma facing components.

PIT VIPER cables are rare-Earth barium copper oxide (REBCO)-based high-temperature superconducting (HTS) cables developed at the Massachusetts Institute of Technology and Commonwealth Fusion Systems. The cable design consists of a twisted copper former with grooves filled with stacks of REBCO tape, enclosed in a structural jacket. The assembly is then filled with solder in a vacuum pressure impregnation (VPI) process. One side effect of the VPI process, however, is potential damage to the superconductor that reduces its critical current and critical exponent. Damage may result from two mechanisms: prolonged exposure to elevated temperatures, and physical erosion of the copper stabilizer layer of the HTS tape, which leaves parts of the REBCO layer unprotected. In this experiment, the effect of extended time scales ($>2$ h) of flowing tin-lead-based solder exposure on HTS in PIT VIPER cables was tested, exploring for the first time exposure to flowing solder at a time scale that is particularly relevant to large-scale magnet manufacturing. During this study, two experimental samples were manufactured for electrical testing: one 2.5-m-long straight cable and one 20-m-long coiled cable. Each cable was exposed to molten tin-lead solder for 2.5 h during the VPI process, and then, electrically tested in a liquid nitrogen bath. Critical current, $n$-value, and resistance were measured. Critical current is compared to modeled values determined from characterization of the tape used, and degradation is assessed from this comparison. The measured critical current, when tested in a liquid nitrogen bath and under self-field, was uniform within reasonable experimental error. This result derisks solder degradation for the manufacturing of SPARC cable magnets.