280

Papers

10/10

Roadblocks Active

3

Connections

FIREFLY is a new software tool that approximates divertor heat loads and particle exhaust efficiency by coupling a simplified stochastic heat transport model to the established EIRENE neutral particle tracker, validated against high-fidelity EMC3-EIRENE simulations on the W7-X stellarator. It enables multivariate optimization of divertor geometry at a fraction of the computational cost of full simulations. Divertor design is one of the most critical and least solved engineering problems for a fusion power plant, and tools that allow rapid iteration directly accelerate the path to viable reactor designs.

██████████ 0.9 divertor-thermal Preprint

ReadSaveConnections

In high-temperature superconducting coils wound from REBCO tape — essential for compact tokamak magnets — the electrical contact resistance between tape layers can drop by up to three orders of magnitude under repeated mechanical cycling at cryogenic temperatures, which undermines quench protection and current sharing. Coating the tape with a 2–3 µm PbSn solder layer or using conductive paste/epoxy largely prevents this degradation, while atomic-layer-deposition and oxidation treatments were less effective. Fusion magnets will experience repeated thermal and mechanical loads over their lifetime, so controlling this resistivity drift is a prerequisite for reliable HTS coil operation.

██████████ 0.9 hts-magnets Preprint

ReadSaveConnections

When neutral gas is injected to stop runaway electrons during a tokamak disruption, the injected atoms recombine with plasma electrons and raise the bulk resistivity; this study uses kinetic modeling and JOREK MHD simulations to show that this resistivity increase selectively amplifies edge tearing modes, making the magnetic field at the plasma boundary stochastic enough to safely disperse the relativistic electrons. The key insight is that the spatial pattern of resistivity — not just its average value — determines which magnetic modes are amplified and where the runaway energy is deposited. This mechanistic understanding explains why certain gas injection strategies produce benign terminations and provides quantitative guidance for disruption mitigation system design on ITER and future devices.

██████████ 0.9 plasma-disruption Preprint

ReadSaveConnections

Using two established bispectral methods (Ritz and Kim) on experimental data from a plasma device, this study shows that energy is transferred nonlinearly from Rayleigh-Taylor instability modes to lower-frequency drift-wave modes through three-wave quadratic coupling — and that whether these analysis methods give reliable results depends critically on the statistical properties of the signal, particularly its kurtosis and spatial stationarity. This matters for fusion because turbulent transport in tokamak edges involves the same classes of instabilities, and knowing when computational turbulence analysis tools are trustworthy is essential before using them to guide machine design or plasma control.

██████████ 0.9 turbulence-modeling Preprint

ReadSaveConnections

A physics-informed neural network (PINN) is trained to solve the drift kinetic equation — which governs how plasma particles travel in a tokamak's non-uniform magnetic field — by embedding the governing equations directly into the training loss function rather than requiring any labeled simulation output data. The resulting surrogate achieves significant computational speedup over conventional numerical solvers and outperforms a data-driven baseline in physical consistency, as tested against EAST tokamak conditions. Faster solutions to this equation enable rapid calculation of neoclassical toroidal viscosity torque, which governs plasma rotation and stability, but the low confidence rating and absent reproducibility information mean these results need independent validation before adoption.

██████████ 0.8 turbulence-modeling Preprint

ReadSaveConnections

A novel pseudo orbit-averaging algorithm combined with phase-space time-dilation achieves a 30,000-fold speedup in gyrokinetic simulations of the WHAM HTS magnetic mirror device (17 T peak field, mirror ratio 32), bridging a nine-orders-of-magnitude gap between fast particle gyration and slow equilibrium timescales. The computed kinetic equilibrium, electrostatic potential, and ion confinement time all agree with analytic Pastukhov confinement theory, providing validation for the method. Magnetic mirrors are experiencing renewed interest as compact fusion candidates, and this capability to directly compute their kinetic plasma equilibrium is a prerequisite for reliable performance predictions.

██████████ 0.8 hts-magnets Preprint

ReadSaveConnections

External magnetic perturbations are routinely used in tokamaks to suppress edge-localized modes (ELMs), but the plasma can shield against these perturbations at resonant surfaces — this paper extends analytic boundary-layer theory to capture ion parallel flow physics that previous models omitted, predicting stronger shielding in parameter regimes relevant to ITER and future devices. The analytic predictions are validated against the GPEC/SLAYER numerical code in a two-fluid drift-MHD framework. Understanding the shielding response quantitatively is important for designing ELM coil systems that actually reach the plasma region they need to perturb.

██████████ 0.8 plasma-disruption Preprint

ReadSaveConnections

Toroidal Alfvén Eigenmodes (TAEs) can be excited by fast alpha particles in a burning fusion plasma and, if they grow too large, scatter those particles away before they transfer their energy to the bulk plasma. Using gyrokinetic particle-in-cell simulations with the ORB5 code, this study shows that for linear drive ratios above about 0.47%, the saturation amplitude is controlled by thermal plasma nonlinearities rather than the fast particle drive, and exhibits a 'stiffness' — meaning doubling the drive barely changes the saturated amplitude. This changes how physicists should think about alpha particle losses in ITER-scale plasmas, where this threshold will likely be crossed routinely.

██████████ 0.8 turbulence-modeling Preprint

ReadSaveConnections

This paper develops an analytical framework that links geometric properties of coarse-grained velocity and magnetic field gradient tensors to the directional flow of energy between spatial scales in MHD turbulence, decomposing contributions from inertial, Maxwell stress, advection, and dynamo mechanisms. Validated with pseudospectral simulations of freely decaying 3D MHD turbulence, the framework provides quantifiable bounds on how much energy flux each mechanism can carry rather than just statistical averages. In fusion plasmas, where MHD turbulence mediates cross-field transport, this type of mechanistic decomposition is needed to build reduced transport models that can be used in whole-device simulations.

██████████ 0.8 turbulence-modeling Preprint

ReadSaveConnections

Right-handed circularly polarized (R-wave) injection into a plasma can suppress runaway electron formation by creating a 'firewall' at the Doppler-shifted cyclotron resonance: particles accelerated by a parallel electric field reach this resonance and receive perpendicular kicks that cancel their parallel energy gain, trapping them below relativistic speeds. The mechanism is demonstrated through classical particle trajectory analysis and particle-in-cell simulations. This identifies wave injection as a potential active mitigation route for runaway electrons during tokamak disruptions — complementary to gas pellet injection — though the practical engineering of such a system at reactor scale remains to be demonstrated.

██████████ 0.8 plasma-disruption Preprint

ReadSaveConnections