Abstract Runaway electron (RE) incidence poses a rather critical threat to the integrity of plasma facing components (PFCs) in tokamaks due to the volumetric nature of energy deposition at relativistic speeds, which leads to steep internal temperature gradients that trigger explosive material failure. A computational framework is presented for the modeling of RE-induced explosive damage in graphite PFCs, benchmarked against a controlled RE impact experiment performed at the DIII-D tokamak. Reconstructed volumetric energy deposition profiles are integrated with high-fidelity three-dimensional thermomechanical simulations based on the finite element method (FEM). It is demonstrated that RE-induced damage in graphite is governed by a degraded subsurface layer formed due to intense volumetric heating, which undermines the structural integrity of the surface and leads to inertial detachment. Continuum thermo-elastic FEM approaches reliably predict the total volume that is liable to detach during failure, while continuum damage-based FEM approaches capture the irreversible subsurface degradation. In contrast to element erosion FEM approaches, a hybrid FEM–to-smoothed-particle-hydrodynamics framework provides physical fidelity, successfully reproducing the experimental blown-off volume and estimating debris kinematics.
Abstract CO 2 is a promising propellant candidate for Hall thrusters due to its cost, storability, and potential for in-situ resource utilization. In this study, fully kinetic particle-in-cell simulations were conducted to elucidate the influence of background pressure in ground test facilities, known as the facility effect, on the performance of a CO 2 Hall thruster. Comparing a no-backpressure condition with dynamic backflow conditions from the chamber demonstrated that utilizing a dynamic inflow model with velocities ranging from tens to hundreds of m s −1 reasonably reproduces the increasing trend of thrust performance observed in experiments. The resulting thrust increments significantly exceeded theoretical values on the assumption that the background gas acts merely as an additional propellant, suggesting that the ingested background gas chain-reactively promotes the ionization of the main propellant via dissociation and multi-step ionization processes. Furthermore, comparisons with simulations where an equivalent gas flow rate was added from upstream indicated that the disproportionate increase in wall ion flux relative to the thrust increment under high backpressure is not unique to downstream backflow. Conversely, structural changes unique to backpressure inflow were indicated. Specifically, the main ionization region is essentially fixed due to the competing ionization effects of the gas inside and outside the channel, and the high background gas density accumulating in the plume region enhances electron mobility from the channel exit to the cathode, relaxing the axial potential gradient. Although the inclusion of dissociation reactions successfully reproduced the decreasing trend in the CO 2 + fraction with increasing backpressure, the underestimation of certain ion beam currents suggests that further model refinement is required. In particular, incorporating charge-exchange collisions and the backflow of dissociation products (CO and O), together with future experiments using varied probe installation distances to evaluate long-path collision effects, will be essential for improving predictive accuracy and validating the model.
Abstract We present the IRE Identification Multimodal Network (IIMN), a classifier to automatically detect Internal Reconnection Events (IREs) in spherical tokamaks. The IIMN utilizes data from the Lithium Tokamak eXperiment-β (LTX-β), a low-aspect-ratio spherical tokamak capable of operating with walls coated in solid or liquid lithium. As a multimodal network, the IIMN is well-suited to handling complex, multi-dimensional relationships among both image and tabular data. To train the model, 7,265 events from 2,968 shots were manually reviewed and labeled, covering nearly two years of diagnostic data from LTX-β. Each event includes key plasma parameters in identifying IREs like the time evolutions of the plasma current, average electron density, and Mirnov spectrogram. The model concatenates the results from a multilayer perceptron branch and a convolutional neural network (CNN) branch to return a one-hot encoded output that determines if the event is an IRE or not based on the event’s plasma parameters. We show that the IIMN has an accuracy of 94.4±0.9%, a precision score of 0.934±0.014, and a recall score of 0.931±0.014. IIMN is fully automated such that it could be used quickly and easily by operators in the control room for spherical tokamak operation. To realize full automation and zero human intervention, our current multimodal architecture incorporates a CNN that directly analyzes image-based data from select plasma diagnostics. By automating IRE detection, this tool can assist operators in monitoring fusion performance and developing strategies to reduce confinement losses even in between two successive plasma discharges.
Abstract Radio frequency (RF) sheath formation at the plasma edge is a critical factor for impurity generation during ion cyclotron range of frequency (ICRF) heating. Motivated by EAST observations that correlate core impurity contamination with edge plasma density and ICRF power, this paper presents a parametric study of RF sheath characteristics using an idealized one-dimensional model. The model couples a hydrodynamic sheath description with an equivalent circuit framework, with the RF driving current prescribed through an RF-drive parameter P RF via a fixed empirical mapping and without antenna-loading feedback. Within the scanned parameter space (edge plasma density n 0 = 0.5 × 10 18 to 1.0 × 10 18 m − 3 and P RF = 1 – 5 MW), spanning the evanescent regime relevant to typical ICRF conditions, the nonlinear balance between displacement and conduction currents yields well-defined scaling laws. The results indicate that, within the scanned density interval, the time-averaged sheath potential scales inversely with the square of the plasma density ( n 0 − 2 ). With respect to P RF , the sheath potential exhibits a stronger-than-classical scaling ( P RF b fit ) where the exponent b fit decreases monotonically from approximately 1.10 at low densities to 0.76 at high densities, approaching the classical P RF 1 / 2 limit. Total harmonic distortion analysis provides a physical interpretation of these scaling behaviors in terms of the density-dependent suppression and P RF -driven enhancement of the displacement current. The implications of these scalings for impurity sputtering are assessed by estimating the physical sputtering yields of metallic impurities (Fe, Cu) under the computed sheath potentials. The physical sputtering yields are shown to be highly sensitive to both parameters. Such results establish the intrinsic parametric dependence of sheath formation within the idealized model, providing theoretical insights into the nonlinear RF sheath effects that can help interpret impurity behaviors in high-power ICRF-heated plasmas.
Simulations of laser-plasma interactions in plasma-based acceleration (PBA) typically require three-dimensional kinetic modeling via the particle-in-cell (PIC) method. While fully explicit 3D PIC codes provide first-principles accuracy, resolving the short laser wavelength(∼ 1µm ) over a long acceleration distance (∼ 100mm ) is computationally expensive. The quasi-static approximation (QSA) mitigates this limitation by decoupling the timescales for the laser evolution and the plasma response, permitting time step for advancing the laser larger than the Courant-Friedrichs-Lewy (CFL) limit. In this article, we implement a new laser envelope solver that uses finite-difference (FD) method in the QSA code QuickPIC. We also couple the solver to a particle-based Ammosov--Delone--Krainov (ADK) field-ionization model for laser-driven ionization. Benchmark simulations against the full-PIC code OSIRIS demonstrate that the new FD-based solver is capable of accurately simulating plasma based accelerator scenarios such as relativistic self-focusing, wakefield excitation, photon acceleration, and field ionization through the laser pulse. Compared with the old laser solver in QuickPIC, the new algorithm is more stable when simulating plasma wakes with large laser amplitude.
