25 June 2024 in Nuclear Fusion
This study systematically explores the parameter space of disruption mitigation through shattered pellet injection in ITER with a focus on runaway electron (RE) dynamics, using the disruption modeling tool Dream. The physics fidelity is considerably increased compared to previous studies, by e.g. using realistic magnetic geometry, resistive wall configuration, thermal quench onset criteria, as well as including additional effects, such as ion transport and enhanced RE transport during the thermal quench. The work aims to provide a fairly comprehensive coverage of experimentally feasible scenarios, considering plasmas representative of both non-activated and high-performance DT operation, different thermal quench onset criteria and transport levels, a wide range of hydrogen and neon quantities injected in one or two stages, and pellets with various characteristic shard sizes. Using a staggered injection scheme, with a pure hydrogen injection preceding a mixed hydrogen-neon injection, we find injection parameters leading to acceptable RE currents in all investigated discharges without activated runaway sources. Dividing the injection into two stages is found to significantly enhance the assimilation and minimize RE generation due to the hot-tail mechanism. However, while a staggered injection outperforms a single stage injection also in cases with radioactive RE sources, no cases with acceptable RE currents are found for a DT-plasma with a plasma current.
19 May 2025 in Journal of Plasma Physics
Accurate modelling of runaway electron generation and losses during tokamak disruptions is crucial for the development of reactor-scale tokamak devices. In this paper, we present a reduced model for runaway electron losses due to flux surface scrape-off caused by the vertical motion of the plasma. The model is made compatible with computationally inexpensive one-dimensional models averaging over a fixed flux-surface geometry, by formulating it as a loss term outside an estimated time-varying minor radius of the last closed flux surface. We then implement this model in the disruption modelling tool DREAM and demonstrate its impact on selected scenarios relevant for ITER. Our results indicate that scrape-off losses may be crucial for making complete runaway avoidance possible even in a $15\,\rm MA$ DT H-mode ITER scenario. The results are however sensitive to the details of the runaway electron generation and phenomena affecting the current density profile, such as the current profile relaxation at the beginning of the disruption.
21 May 2025 in Journal of Plasma Physics
Electron energisation by magnetic reconnection has historically been studied in the Lagrangian guiding-centre framework. Insights from such studies include that Fermi acceleration in magnetic islands can accelerate electrons to high energies. An alternative Eulerian fluid formulation of electron energisation was recently used to study electron energisation during magnetic reconnection in the absence of magnetic islands. Here, we use particle-in-cell simulations to compare the Eulerian and Lagrangian models of electron energisation in a set-up where reconnection leads to magnetic island formation. We find the largest energisation at the edges of magnetic islands. There, energisation related to the diamagnetic drift dominates in the Eulerian model, while the Fermi related term dominates in the Lagrangian model. The models predict significantly different energisation rates locally. A better agreement is found after integrating over the simulation domain. We show that strong magnetic curvature can break the magnetic moment conservation assumed by the Lagrangian model, leading to erroneous results. The Eulerian fluid model is a complete fluid description and accurately models bulk energisation. However, local measurements of its constituent energisation terms need not reflect locations where plasma is heated or accelerated. The Lagrangian guiding centre model can accurately describe the energisation of particles, but it cannot describe the evolution of the fluid energy. We conclude that while both models can be valid, they describe two fundamentally different quantities, and care should be taken when choosing which model to use.
27 August 2025 in Nuclear Fusion
A couple of technical issues with our simulations presented in Vallhagen et al (2024 Nucl. Fusion 64 086033) have been corrected, which had a quantitative effect on the simulations in the original paper. This corrigendum summarizes these corrections and provides updated simulation results.
11 June 2026 in Journal of Plasma Physics
Bridging the spatiotemporal scales of magnetic seed field generation and subsequent dynamo amplification in the weakly collisional intracluster medium presents an extreme numerical challenge. We perform collisionless turbulence simulations with initially unmagnetised electrons that capture both magnetic seed generation via the electron Weibel instability and the ensuing dynamo amplification. Going beyond existing pair-plasma studies, we use an ion-to-electron mass ratio of $100$ for which we find electron and ion dynamics are sufficiently decoupled. These simulations are enabled by the 10-moment collisionless fluid solver of Gkeyll , which evolves the full pressure tensor for all species. The electron heat-flux closure regulates pressure isotropisation and effectively sets the magnetic Reynolds number. We investigate how the strength of the closure influences the transition between a regime reminiscent of previous kinetic pair-plasma simulations and a regime exhibiting dynamo behaviour qualitatively similar to magnetohydrodynamics.