Bulletin of the American Physical Society
61st Annual Meeting of the APS Division of Plasma Physics
Volume 64, Number 11
Monday–Friday, October 21–25, 2019; Fort Lauderdale, Florida
Session GI2: Invited MF: Stellarators and Computational Techniques |
Hide Abstracts |
Chair: Benjamin Faber, University of Wisconsin Room: Floridian Ballroom AB |
Tuesday, October 22, 2019 9:30AM - 10:00AM |
GI2.00001: Observation of isotope separating and mixing states in isotope mixture plasmas in LHD Invited Speaker: Katsumi Ida Ion and electron density profiles are identical due to quasi-neutrality in single isotope plasmas. However, in mixed isotope plasmas, individual isotope density profiles can differ. Measurements of radial hydrogen (H) and deuterium (D) density profiles, using bulk charge exchange recombination spectroscopy, reveal two states of isotope particle transport in H-D in LHD plasmas[1]. One is isotope-separating (IS), where the source location impacts the density ratio profile and the other is isotope-mixing (IM) where the density ratio profile is flat regardless of the location of the H and D source. Gyrokinetic simulations predict an IM state in ITG turbulence regimes and an IS state in TEM regimes[2]. The IS state is observed in the low-density plasmas ($n_e$ $\sim$ 1.5 $\times$ 10$^{19}$m$^{-3}$, $dn_e/dr$(0.8$\rho$) $<$ 0) where the beam fueling isotope species differ from the isotope species due to recycling. The H/(H+D) density profile is peaked for H beam fueling and D recycling wall conditions. A peaking of D/(H+D) density is observed for D beam fueling and H recycling wall conditions. When the wall recycling is equally mixed, (H/D $\sim$ 1), H/(H+D) dominates and D/(H+D) disappears. In contrast, the IM state is observed in higher density plasmas ($n_e$ $\sim$ 3$\times$ 10$^{19}$m$^{-3}$, $dn_e/dr$ (0.8$\rho$) $>$ 0) with shallow pellet injection. H and D pellets have been deposited at 0.9$\rho$ in the plasmas with peaked H/(H+D) density profile. This results in the H/(H+D) profile becoming flat after the H and D pellets are injected, which clearly results in the IM state. These results demonstrate that either of two isotope states (IS or IM) can exist in mixed H and D plasmas depending on collisionality and density gradient and provides important insight into the control of isotope density ratio profiles needed for DT operations in tokamaks. [1] K. Ida, et. al., Nucl. Fusion 59 (2019) 056029. [2] C. Bourdelle et. al., Nucl. Fusion 58 (2018) 076028. [Preview Abstract] |
Tuesday, October 22, 2019 10:00AM - 10:30AM |
GI2.00002: Evidence of Turbulence-Induced Ion Temperature Limitations In Wendelstein 7-X Invited Speaker: Sergey Bozhenkov The stellarator Wendelstein 7-X is optimized for reduced neoclassical transport, allowing access to high-performance plasmas at high ion temperatures. Indeed, ion temperatures above 3 keV and energy confinement times above the ISS04-scaling were transiently achieved in post-pellet phases, from which it can be preliminary concluded that the optimization is effective. However, in the vast majority of discharges, the plasma performance is below neoclassical predictions, and the energy confinement time is at the level of the ISS04-scaling. For these discharges, Ti, in contrast to Te, is limited below 2 keV, suggesting a high degree of ion temperature profile stiffness. No dependence on the efficiency of collisional coupling between ions and electrons is observed. The addition of direct ion heating by NBI to such plasmas does not in general improve the situation. Also, no significant dependence on the magnetic configuration is found.\\ During the post-pellet high-performance phases, a clear reduction of density fluctuations is observed, strongly suggesting that turbulence stabilization is responsible for the higher confinement. This hypothesis is supported by GENE simulations, showing that the linear growth rate and nonlinear ion heat flux are minimized when the normalized ion temperature and density gradients are of similar magnitude and overlap. The effect is caused by a twofold stabilizing influence of trapped electrons in maximum-J configurations, enhanced ITG suppression, as well as by a relatively weak TEM response.\\ These findings demonstrate that the plasma performance in W7-X is to a large degree determined by turbulent losses. It is, therefore, important to study different methods and scenarios for turbulence reduction, e.g. H-mode-like regimes, or turbulence-optimized configurations. [Preview Abstract] |
Tuesday, October 22, 2019 10:30AM - 11:00AM |
GI2.00003: Adjoint methods for efficient stellarator optimization and sensitivity analysis Invited Speaker: Elizabeth Paul The design of modern stellarators often employs gradient-based optimization to navigate the high-dimensional spaces used to describe their geometry. However, computing the numerical gradient of a target function with respect to many parameters can be expensive. The adjoint method allows these gradients to be computed at much lower computational cost and without the noise associated with finite differences. This technique has been employed widely in automotive and aerodynamic engineering, and we present the first applications to stellarator design. An adjoint method has been implemented in the stellarator coil design code REGCOIL, allowing for optimization of the coil-winding surface with analytic gradients to obtain improved coil shapes with minimal field error. An adjoint drift kinetic equation has also been implemented in the SFINCS code to compute gradients of moments of the distribution function, such as the bootstrap current, with respect to geometric parameters. We apply this method for optimization of neoclassical quantities without inherent assumptions on the collisionality or radial electric field. Furthermore, we present an adjoint method for obtaining the gradients of functions of MHD equilibria, such as the rotational transform or magnetic well, with respect to the shape of the plasma boundary or coils, providing an order $10^2-10^3$ reduction in cost. In addition to gradient-based optimization, we use the derivatives obtained from the adjoint method for local sensitivity analysis. For example, derivatives from REGCOIL inform where the normal field error is most sensitive to displacements of coils. Similarly, derivatives computed from SFINCS inform where the bootstrap current is most sensitive to perturbations of the field strength. These local sensitivity calculations provide quantification of engineering tolerances and insight into optimization. [Preview Abstract] |
Tuesday, October 22, 2019 11:00AM - 11:30AM |
GI2.00004: Competition between parallel viscosity and ion-neutral friction in damping the parallel flow in a quasisymmetric stellarator Invited Speaker: Santhosh Kumar Experimentally measured parallel ion flows and radial electric fields in the quasisymmetric configuration of the HSX stellarator have been shown previously to be inconsistent with standard neoclassical calculations. To examine this inconsistency, improvements have been made in both experimental technique and theoretical modeling. A charge exchange spectroscopy system has been upgraded to measure the counter-streaming Pfirsch-Schl\"{u}ter parallel ion flows from which flux surface averaged parallel ion flows and the radial electric field can be obtained without using the radial force balance equation. This method provides an improved measurement of the parallel ion flows and radial electric fields in the core of the plasma where the poloidal flow measurements have large uncertainties. Along with the improvements in diagnostics, the neoclassical transport code PENTA has been modified to include collisions with background neutrals. Toroidal and poloidal arrays of H-alpha detectors in conjunction with the neutral transport code DEGAS provided radial and temporal profiles of neutral density in the plasma. Including neutral friction in the neoclassical calculation significantly damps the parallel ion flow and resolves the inconsistency of the parallel flow and radial electric field for the quasisymmetric magnetic geometry. The calculated radial electric field values are relatively unchanged with the inclusion of neutral friction. As the decrease in ion parallel flow is matched by an increase in the electron flow, the bootstrap current is also found to be insensitive to the neutral friction. In a configuration of HSX in which the quasisymmetry is intentionally degraded, the effect of neutral friction is found to be less important due to the increase in the neoclassical parallel viscosity. [Preview Abstract] |
Tuesday, October 22, 2019 11:30AM - 12:00PM |
GI2.00005: Elimination of alpha particle losses in quasi-helically symmetric stellarators Invited Speaker: Aaron Bader In three-dimensional equilibria non-zero bounce-averaged radial drifts may exist. Radial drifts can cause losses of both thermal and energetic particles. One possible method of eliminating drifts is through quasi-symmetry, as a perfectly quasi-symmetric surface will have no bounce-averaged radial drift. However, in realistic stellarator configurations there will always be deviations from quasi-symmetry, and these deviations will drive energetic particle losses. Previous calculations of stellarator equilibria showed that some collisionless alpha particle losses always existed even deep within the core. In this talk we demonstrate that there exist configurations with no energetic particle losses in the core plasma, and losses well below 1\% within the mid-radius. The mechanism for producing these configurations was non-linear equilibrium optimization for quasi-helical symmetry and a metric developed by V.V. Nemov, $\Gamma_c$ [1], that seeks to align the second adiabatic invariant, $J$ with the flux surfaces. We demonstrate that the methodology succeeds in greatly reducing losses of particles near the trapped-passing boundary, where most losses are concentrated. [1] V.V. Nemov PoP (2008) 15, 052501 [Preview Abstract] |
Tuesday, October 22, 2019 12:00PM - 12:30PM |
GI2.00006: First Implementation of Gyrokinetic Exact Linearized Landau Collision Operator and Comparison with Models Invited Speaker: Qingjiang Pan The gyrokinetic exact Landau operator has now been formulated in conservative and symmetric Landau form\footnote{Q. Pan and D. R. Ernst, Phy. Rev. E \textbf{99}, 023201 (2019)} and implemented for the first time in a gyrokinetic code (GENE). The new exact operator makes it possible to assess the accuracy of widely used model collision operators for the first time. The gyrokinetic Landau form, though equivalent to the Rosenbluth form\footnote{B. Li and D. R. Ernst, Phys. Rev. Lett. \textbf{106}, 195002 (2011)}, explicitly preserves the symmetry between test- and field-particle terms. This symmetry underlies the conservation laws and the H-theorem, and enables finite-volume or spectral methods to preserve the conservation, independent of resolution. The present implementation utilizes the same finite-volume method recently employed to discretize the Sugama collision model in GENE\footnote{P. Crandall et al., submitted to Comput. Phys. Commun. (2018)}, allowing direct comparison between the two operators. Neoclassical tests confirm that the Sugama model overestimates ion heat fluxes by about 20-25\% relative to the exact operator. The exact operator has now been used in nonlinear gyrokinetic simulations of density-gradient-driven trapped electron mode (TEM) turbulence with finite Larmor radius (FLR) collisional corrections. Results are (1) the Sugama model operator underestimates TEM growth rates by about 10\%; (2) the exact operator and the Sugama model produce similar particle and heat fluxes near the nonlinear threshold; and (3) the Sugama model yields accurate zonal flow and GAM damping rates. The finite-volume scheme, though conservative, requires high velocity resolution for convergence. Future work will improve the performance by using spectral methods and implicit time stepping. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700