277

Papers

10/10

Roadblocks Active

2

Connections

Simulations of ITER- and SPARC-scale burning plasmas find that alpha particles — the helium nuclei produced by fusion reactions — mildly excite a class of magnetic wave called toroidal Alfvén eigenmodes, which in turn amplify rotating flow structures (zonal flows) that shear apart and suppress the dominant source of turbulent heat loss. The mechanism is nonlinear and self-regulating: more fusion power produces more alphas, which produce stronger zonal flows, which reduce energy losses. If confirmed experimentally, this means burning plasma confinement may be intrinsically better than models trained on non-burning plasmas predict.

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

ReadSaveConnections

This paper introduces PT_SOLVER, a new numerical module inside the widely-used TRANSP tokamak modeling code, designed to handle the extreme numerical stiffness that arises when turbulent transport depends sensitively on temperature and density gradients. The module uses an implicit Newton method with stabilization tailored to steep-gradient regions, verified against analytical benchmarks and independent codes. Better numerical robustness in transport solvers directly reduces uncertainty in design predictions for ITER and future reactors, where incorrect handling of stiff transport can cause simulations to predict wildly different plasma profiles.

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

ReadSaveConnections

This study uses multi-dimensional finite-element simulations to show that standard one-dimensional models significantly misinterpret hydrogen isotope permeability measurements in molten FLiBe salt (a candidate tritium-breeding coolant) at 773–973 K, because lateral leakage pathways are ignored. The true permeability inferred from 2D/3D models shows Arrhenius temperature dependence but with an absolute value that shifts considerably depending on boundary conditions at vessel walls. Accurate tritium permeability data is essential for predicting how much tritium leaks into or out of the breeding blanket — a safety and fuel-cycle concern for any fusion power plant using molten salt.

██████████ 0.8 tritium-breeding Preprint

ReadSaveConnections

Molecular dynamics simulations reveal that helium is nearly insoluble in liquid lead-lithium alloy (a leading tritium-breeding blanket fluid), causing it to form bubbles governed by strong interfacial tension and local pressure gradients at the liquid-metal interface. This matters because helium is produced inside the blanket by neutron reactions, and uncontrolled bubble formation could disrupt tritium extraction efficiency and cause structural stress on blanket walls. The results provide atomistic parameters — interfacial tension vs. temperature — that are needed as inputs to engineering-scale blanket flow models.

██████████ 0.8 tritium-breeding Preprint

ReadSaveConnections

This paper demonstrates a statistical framework that can certify the safety of fusion plasma-facing components (such as the tungsten tiles lining the divertor) using computer models alone, without requiring every design variant to undergo expensive physical testing at reactor-relevant heat loads above 10 MW/m². The approach carefully separates model errors from measurement uncertainty using a modified area validation metric, and the authors report that benchmark datasets are published. Reducing the experimental testing burden for component qualification is practically important because ITER-scale heat flux facilities are few, costly, and time-limited.

██████████ 0.8 divertor-thermal Preprint

ReadSaveConnections

This paper reports the first-ever construction and test of a high-temperature superconductor (HTS) magnet wound in a canted-cosine-theta geometry using ReBCO tape, operated at ~46.5 K without liquid cryogens. Although the target application is a particle collider, the manufacturing and qualification techniques — winding ReBCO tape on precision-machined formers, paraffin wax impregnation, cryocooler testing — are directly transferable to the compact high-field tokamak magnets (such as those in SPARC) that require HTS to achieve high Q. The single-prototype nature and IP-protected winding trajectory are current limitations.

██████████ 0.8 hts-magnets Preprint

ReadSaveConnections

This theoretical work derives Kolmogorov's classical turbulence scaling law (which predicts how energy distributes across eddy sizes) as the solution to a variational minimization problem in a free-energy framework, requiring specific conditions on the energy cascade called Downhill Flux Conditions. For fusion, turbulent transport in tokamaks is ultimately governed by analogous cascade physics, and improved mathematical foundations for turbulence theory can sharpen the predictive models used in plasma transport codes. The result is formal and analytical; direct application to magnetized plasma turbulence would require further development.

██████████ 0.8 turbulence-modeling 🔗 4 cited Peer-reviewed

Read

The Jenga integrated design framework is validated on the MAST-U spherical tokamak by reproducing plasma equilibrium shapes (flux surfaces) from experimental data across multiple operating scenarios — including the advanced Super-X divertor — and agrees with two independent codes. The framework also couples equilibrium results to neutronics calculations using OpenMC, enabling full-system analysis from a single workflow. Integrated frameworks that connect plasma physics to materials and neutronics calculations are essential for designing the breeding blankets and structural components of future fusion plants.

██████████ 0.7 first-wall-materials Preprint

ReadSaveConnections

This paper provides a systematic mathematical derivation showing that the Boris algorithm — the standard method used in plasma simulation codes to advance charged particles through magnetic fields — is a special case of a broader class of exponential integrators, and that a related Rodrigues scheme is theoretically more accurate when field values are recalculated at each time step. Understanding the hierarchy of particle integration methods matters for simulating fast alpha particles and runaway electrons in turbulent tokamak fields, where numerical energy conservation errors can accumulate over millions of time steps.

██████████ 0.7 turbulence-modeling Preprint

ReadSaveConnections

This paper upgrades the HPI2 pellet ablation code by deriving a physically consistent numerical step size from ablation physics, replacing a hand-tuned parameter, and validates the improved model against a WEST tokamak discharge with ~10% error in predicted density increments. The upgrade is then embedded in a full predictive integrated modeling workflow (JINTRAC/IMAS) that correctly reproduces density rise and electron temperature transients after pellet injection. Pellet injection is a primary method for fueling the dense plasma cores needed in reactor-grade tokamaks, so accurate predictive models reduce the experimental iteration needed to optimize fueling strategies.

██████████ 0.7 long-confinement Preprint

ReadSaveConnections