Speaker
Description
The physical mechanism behind the difference in the H-mode power threshold between favourable and unfavourable drift configurations is still not fully understood and remains an active field of research in view of ITER operation. Experimental evidence indicates that the radial electric field (Er) shear facilitates H-mode access in the favourable configuration, while it is weaker in the unfavourable case [1]. However, the role of turbulence in the self-consistent interplay between transport, equilibrium profiles, and Er has not yet been clarified.
Here, we leverage the recent computational boost in the full-f gyrokinetic code GENE-X [2–4] to perform first-of-their-kind edge and scrape-off layer (SOL) turbulence simulations with X-points of a matched pair of L-mode ASDEX Upgrade discharges in favourable and unfavourable drift configurations [1]. The simulations are validated against experimental measurements and show excellent agreement. In particular, the experimental Er profiles are reproduced, and the simulations demonstrate that poloidal flows provide the dominant contribution to the radial force balance. Most importantly, while these simulated flows remain close to the neoclassical level in the unfavourable configuration, they significantly exceed neoclassical predictions in the favourable case, producing the experimentally observed stronger Er shear. We identify turbulence-driven poloidal flows [5], arising from enhanced nonlinear turbulence–mean-flow energy transfer driven by the radial gradient of the total velocity stress (a full-f generalization of the Reynolds stress [6]), as the key mechanism underlying the observed difference in Er between favourable and unfavourable configurations. This nonlinear energy transfer is much weaker in the unfavourable configuration, yielding a shallower Er well. Edge turbulence and geodesic acoustic mode activity are also found to be weaker in the favourable configuration. The universality of these findings is demonstrated by extending the present analysis to the TCV tokamak. Within the turbulence–flow shear suppression paradigm, the deeper Er well and reduced turbulence intensity in the favourable configuration may underlie the observed configuration dependence of the H-mode power threshold. Extensions toward H-mode are presented by attempting a sudden increase in the input power [7].
References
[1] U. Plank et al., Phys. Plasmas 30, 042513 (2023)
[2] D. Michels et al., Comput. Phys. Comm. 264, 107986 (2021)
[3] B. J. Frei et al., Comput. Phys. Comm. 316, 109817 (2025)
[4] B. J. Frei et al., Nucl. Fusion 65, 116026 (2025)
[5] B. J. Frei et al., Submitted to Phys. Rev. Lett. (2026)
[6] O. Grover et al., Nucl. Fusion 64, 056020 (2024)
[7] W. Zholobenko et al., Phys. Rev. Lett. (2026)
*Correspondence email: baptiste.frei@ipp.mpg.de