Description
Fusion reactors will have significantly smaller normalized gyroradius $\rho_* = \rho_i/a$ than present experiments. This fact has a rigorous mathematical consequence: if pedestal and edge gradients are bounded by gyrokinetic threshold and transport constraints [parisi:2024], then as $\rho_* \rightarrow 0$, the region governed by non-gyrokinetic physics (orbit loss-cone, etc.) becomes asymptotically thin, so that the gyrokinetic-governed volume dominates the confined plasma. Moreover, without neutral-beam torque or edge-localized heating, reactor edge confinement may take the form of a mild, I-mode-like shoulder rather than a steep, MHD-limited pedestal, further extending the gyrokinetic domain. Together, these considerations motivate a shift in perspective: rather than viewing gyrokinetics as marginal or inapplicable at the edge, future devices may realize a gyrokinetic edge in which drift-wave turbulence, modified by moderate ExB shear, sets transport all the way to within a few gyroradii of the separatrix.
In this work we systematically apply CGYRO [candy:2016] to reactor-edge conditions, treating the separatrix temperature and density as boundary conditions set by SOL/divertor physics. Recent global spectral gyrokinetic results [candy:2025] indicate that nonlocal effects are largely trivial, with the principal exception of a shear-like contribution arising from profile variation; this suggests that standard ExB shear can be systematically incorporated into local edge simulations to mimic the effect of profile shear. The focus is on mapping the boundary between well-behaved turbulence regimes and nonstandard or unbounded kinetic turbulence as the edge orbit-loss region is approached. This framing positions gyrokinetics as the natural theoretical description of the reactor edge, where reduced reliance on non-gyrokinetic pedestal models is warranted.
This work was supported by the U.S.\ Department of Energy under awards DE-FG02-95ER54309 and DE-SC0024425 (FRONTIERS SciDAC-5 project); and used computer resources of the OLCF under Contract DE-AC05-00OR22725, the ALCF under contract DE-AC02-06CH11357 and NERSC under Contract DE-AC02-05CH11231.
[parisi:2024] Parisi J et al 2024 Nucl. Fusion 64 086034
[candy:2016] Candy J, Belli E and Bravenec R 2016 J. Comput. Phys. 324 73
[candy:2025] Candy J, Dudkovskaia T and Belli E 2025 Phys. Rev. E 111 L053201