Orbit modeling of fast particle redistribution induced by sawtooth instability

Abstract

Initial tests on the National Spherical Torus Experiment Upgrade (NSTX-U) (Menard et al 2012 Nucl. Fusion 083015) device suggest that introducing energy selectivity for sawtooth induced fast ion redistribution is required to improve the agreement between experimental and simulated quantities such as neutron rate and Fast-Ion D-Alpha profiles. The aim of this work is to assess the requirements to properly describe the behaviour of fast ions during a sawtooth crash for predictive sawtooth simulations. As the first step, in this work, we use the particle-following Orbit code to characterize the redistribution of fast particles. In order for a sawtooth crash to be simulated, a spatial and temporal displacement is implemented into the Orbit code. The perturbation amplitude is determined by comparison with experimental measurement of the neutron rate drop. The characteristics of fast ions with different orbit types are investigated in phase and real space. Due to a sawtooth crash, fast ion energy and angular momentum are modified resulting in the redistribution in phase space and orbit type change. The redistribution of fast ions in real space shows that the sawtooth instability brings different effect on fast particles with different orbit types as observed in experiments. The initial interpretative Transp simulation using the so-called kick model based on the Orbit modeling result shows an improvement of fast ion redistribution before and after a sawtooth crash but the neutron rate still has discrepancy compared to the experimental measurement.

Simulating energetic particle losses in JET plasmas with a reverse integration biasing scheme

“…The radial displacement is represented as the dot product of the plasma displacement vector, ξ, with the gradient of the poloidal flux, ∇ψ p , and is plotted across the radial position defined by the poloidal flux normalized to the flux at the last closed flux surface, ψ p /ψ p (a). The mode structure is in agreement with analytic kink-like representations [43,45] and results from the KINX code [46] which exhibit a dominant, core-localized, m = 1 component with subdominant m = 2, 3 components extending radially outward.…”

“…Since the n = 1 kink mode appears to drive the largest measureable losses, the other toroidal mode numbers will be ignored for simplicity [43]. In practice, however, the model can contain multiple modes.…”

Numerical studies on saturated kink and sawtooth induced fast ion transport in JET ITER-like plasmas

“…Potato/stagnation orbits undergo the most changes since the topological boundaries are relatively close together to other orbit types. These results are similar to those in NSTX-U sawtooth modeling and experiment which exhibited comparable orbit transitions but did not observe any redistribution above the critical energy range at high pitch [10,34]. In short, the reduced sawtooth modeling has revealed strong resonant transport effects which are typically not included in modeling efforts such as the standard Kadomtsev model within TRANSP.…”

“…specific areas of pitch and gyroradius). in gyroradius while the high energy PMT corresponds to the region of ∼[45, 55] • and ∼ [10][11][12][13] cm, respectively. It is important to note that the PMT array does not fully cover the CCD plate domain, so the total integrated CCD signal contains additional losses not accounted for among the PMTs.…”

“…Internal kink modes and sawteeth have previously been reported to alter the fast ion distribution and produce measureable losses [2][3][4][5][6]. The mechanisms of transport differ between resonant [7][8][9] and non-resonant [10,11] processes for the two instabilities. The kink and sawtooth crash have preferences for mode-resonant or energetic particle phase-space redistribution/stochastization induced transport.…”