Description
For the development of operating scenarios for future tokamaks, like ITER, it is essential that anomalous plasma transport can be efficiently and accurately modeled. Such modeling capabilities must be available for the full plasma radius and all phases of a discharge, including ramp-up and ramp-down. In this work, we assess the ability of three reduced transport models: (i) the semi-empirical Bohm-gyroBohm (BgB) model[1], (ii) the EDWM model[2] and (iii) TGLF-SAT2[3,4] with electrostatic settings, to simulate temperature and density profiles from the flat top phase of two ASDEX Upgrade (AUG) L-mode plasmas. Of (i)-(iii), TGLF-SAT2 is expected to be the most accurate; recent work indicate that it can be used in integrated modeling to reproduce the dependency of stored energy on engineering parameters seen in scaling laws[5], and to successfully predict kinetic profiles in L-mode tokamak plasmas[6]. Simulations in this work have been performed with the European Transport Simulator (ETS)[7], an integrated modeling framework which couples multiple physics models to make predictions and interpretations of kinetic plasma profiles in tokamak experiments.
ETS predicted density and temperature profiles indicate that transport models (i)-(iii) are least accurate for plasmas that are purely EC heated, with overpredictions of central electron temperatures and underpredictions of ion temperature in the same region. BgB and TGLF-SAT2 predict total thermal energy content that is within 15% of experimental values regardless of auxiliary heating mechanism. EDWM does not capture electron scale turbulence, and therefore is the least accurate of the three models. Eigenvalue spectrums from standalone TGLF-SAT2 ky-scans on ETS predicted and experimental data compared to linear GENE[8] on experimental data indicate that the two codes qualitatively agree on the most unstable modes on the ion scale, but diverge for higher ky. Overall, results from this work indicate that ETS with BgB and TGLF-SAT2 can predict total thermal energy content within 15% of experimental values, yielding relative errors in density and temperature profiles that range from <10% (best) to >60% (worst). Further investigation is needed into modeling purely EC heated plasmas, as predictions get less accurate under these circumstances.
[1] V.V. Parail et al. Nucl. Fusion, 39(3):429, 1999
[2] P. Strand et al. In Proceedings of the 31st EPS Conference
on Plasma Physics, pages 28G P–5.187, 2004
[3] G. M. Staebler et al. Phys. Plasmas, 14(5):055909, 2007.
[4] G. M. Staebler and J. E. Kinsey. Phys. Plasmas, 17(12):
122309, 201
[5] C. Angioni et al, Nucl. Fusion, 62 066015, 2022
[6] G.M Staebler et al, Nucl. Fusion 64 085002, 2024
[7] D. P. Coster et al. IEEE Trans. on Plasma Science, 38:2085–2092, 2010
[8] F. Jenko et al. Phys. Plasmas, 7(5):1904–1910, 2000