285

Papers

10/10

Roadblocks Active

4

Connections

This study shows that a 'second stability region' — where plasma remains stable against turbulence even at high pressure gradients — exists in gyrokinetic theory and is entered by tokamak plasma in H-mode because bootstrap current and the Shafranov shift reshape the magnetic geometry. Entry requires a normalized pressure gradient above roughly 1 at low collisionality, conditions that H-mode routinely satisfies. This gives a concrete gyrokinetic mechanism for why H-mode confinement is improved: the plasma crosses into a regime where the dominant turbulence drive is geometrically suppressed, supporting higher fusion power density.

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

ReadSaveConnections

Using gyrokinetic theory benchmarked against the GENE code, this paper derives compact analytical expressions showing that a higher fraction of fast ions (e.g., fusion-born alpha particles) monotonically raises the critical temperature gradient at which ion temperature gradient (ITG) turbulence — the dominant heat loss mechanism in tokamak cores — switches on. The physical mechanism is that fast ions, with gyroradii larger than thermal ions, dilute the drive for ITG while modifying the effective charge profile. For burning plasmas where alpha heating dominates, this self-stabilizing effect could reduce turbulent transport below what cold-plasma models predict, materially affecting Q predictions.

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

ReadSaveConnections

This review synthesizes how physics-informed neural networks and Fourier neural operators can be coupled with conventional MHD solvers to simulate plasma turbulence at fusion-relevant Reynolds and Lundquist numbers (exceeding 10^10) that are completely intractable for direct numerical simulation alone — the required degrees of freedom scale as Re^(9/4) or faster. The core argument is that AI trained on leadership-scale simulations can extrapolate into regimes beyond classical discretization, offering a credible route to predictive turbulence modeling without waiting for exascale hardware. As a survey rather than original research, its findings carry low confidence individually but map the methodological frontier.

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

ReadSaveConnections

An experiment on the IMPED linear plasma device systematically varied the magnetic field from 600 to 1000 G and found that stronger fields suppress large-scale zonal flows while amplifying mean flows, shifting turbulent power from low frequencies (0.1–1 kHz) to high frequencies (1–300 kHz). This demonstrates directly that magnetic field strength does not merely confine plasma but actively redistributes energy across turbulence scales — relevant to understanding why edge transport changes with toroidal field in tokamaks. The result is quantified through Reynolds stress measurements and spectral slope changes, providing clean experimental benchmarks for turbulence codes.

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

ReadSaveConnections

Using OpenMC neutron transport simulations, this paper shows that 14 MeV D-T fusion neutrons can produce radioisotopes for nuclear batteries — including Pm-147, Ni-63, Ar-39, and Cs-137 — at rates orders of magnitude higher than current fission-reactor methods, while the same blanket feedstock simultaneously breeds tritium to close the fuel cycle. The dual-use concept is significant for fusion economics: producing high-value isotopes as a byproduct could offset plant operating costs before electricity generation becomes the primary revenue source. The analysis is computational and prospective, using simplified slab geometry as the primary case.

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

ReadSaveConnections

This simulation study shows that the polarization of magnetic perturbations — whether Alfvénic or magnetosonic — strongly controls how far magnetic field lines wander from their nominal path, which in turn sets how fast heat and particles escape across a confinement device. A key finding is that magnetosonic-polarized turbulence changes the asymptotic field-line diffusion scaling from (δB/B)² to (δB/B)⁴, meaning the turbulence character, not just its amplitude, determines the transport level. This has direct implications for modeling anomalous heat diffusivity in tokamak edge and scrape-off-layer regions where different wave modes coexist.

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

ReadSaveConnections

This study benchmarks three algorithms — Minimum Fisher Information, Phillips-Tikhonov Regularization, and MLEM — for reconstructing 2D plasma radiation emissivity maps from the line-integrated images of an infrared bolometer camera on the ADITYA-U tokamak, using four synthetic phantoms representing realistic emission profiles. Accurate radiation maps are critical for managing divertor power loads and detecting impurity accumulation in real time. The algorithms show distinct tradeoffs: peak preservation, noise robustness, and convergence speed differ significantly, informing which method is best suited for real-time control applications versus offline analysis.

██████████ 0.7 divertor-thermal Preprint

ReadSaveConnections

The NESSA facility at Uppsala has been commissioned as a compact D-T neutron source achieving a calibrated yield of 4.7×10⁸ neutrons per second at 14.1 MeV, verified using Nb-93 activation foils and cross-validated against four independent Monte Carlo codes (FLUKA, Geant4, MCNP, PHITS). Such facilities fill a critical gap: testing candidate first-wall and breeding-blanket materials under fast-neutron conditions without needing a full fusion device. The agreement between four simulation codes provides confidence in using this source for quantitative materials qualification studies.

██████████ 0.6 first-wall-materials Preprint

ReadSaveConnections

This paper derives and numerically solves the kinetic equation governing turbulence among fast magnetosonic waves in low-β plasmas, finding a Kolmogorov-Zakharov energy spectrum scaling as k^(-3/2) and a mixed energy cascade where co-propagating waves transfer energy backward (to large scales) while counter-propagating waves cascade forward. Fast magnetosonic waves are directly relevant to ion cyclotron range of frequency (ICRF) heating schemes used in tokamaks, and understanding their turbulent energy redistribution informs how efficiently heating power reaches the plasma core rather than being scattered near the antenna.

██████████ 0.6 turbulence-modeling Preprint

ReadSaveConnections

Two-dimensional MHD simulations of coalescing magnetic islands show that once reconnection enters the plasmoid-mediated regime — expected at the high Lundquist numbers of fusion plasmas — the reconnection rate becomes nearly independent of the magnetic Prandtl number (the ratio of viscosity to resistivity). This matters for disruption physics: tokamak disruptions involve rapid reconnection events, and this result suggests the reconnection rate is robustly fast in the plasmoid regime regardless of viscosity, supporting the picture that disruptions, once initiated, proceed rapidly and are difficult to slow by changing plasma viscosity.

██████████ 0.5 plasma-disruption Preprint

ReadSaveConnections