Bulletin of the American Physical Society
63rd Annual Meeting of the APS Division of Plasma Physics
Volume 66, Number 13
Monday–Friday, November 8–12, 2021; Pittsburgh, PA
Session PO09: MFE: KSTAR and IgnitorOn Demand
|
Hide Abstracts |
Room: Rooms 403-405 |
Wednesday, November 10, 2021 2:00PM - 2:12PM |
PO09.00001: Overview of KSTAR Si-Woo Yoon, J. Chung, W. H. Ko, G. Shin, H. S. Hahn, Y. M. Jeon, J. H. Lee, M. J. Choi, Yong-Su Na, Jin Myung Park, Steven A Sabbagh The recent progress of the KSTAR program is summarized. First of all, continuous effort has been made to develop new advanced scenarios targeting steady-state operation with high beta sustainment which includes development of the hybrid and high li-mode aided with control-oriented modeling on transport. In addition, the stationary internal transport barrier is successfully reproduced with comparable confinement as H-mode in high power neutral beam injection with I-mode like edge configuration and the technical issues are identified for long-pulse sustainment of the advanced scenarios. Recent experiments on resonant magnetic perturbation (RMP) have focused on the controllability of ELMs in real-time, such as suppression of the first ELM right after H-mode access with real-time controller, the adaptive control of ELM suppression with minimum confinement loss from RMP application, and the effect of up-down asymmetry of the applied RMP on ELM suppression window which are facilitated with the new control algorithm based on the Machine Learning (ML). Turbulence in pedestal is further investigated for its role of the ELM suppression and the correlation of edge turbulence with RMP hysteresis and ion pedestal broadening are also investigated. In addition, stochastic fluctuation and its impact on the transport in the edge region is also identified with a detailed analysis of ECEI diagnostic data. Together with the experimental validation of the multiple Shattered Pellet Injections (SPIs), an innovative model of the disruption event is developed for characterization and forecasting based on the real-time application of plasma control system and diagnostics. |
Wednesday, November 10, 2021 2:12PM - 2:24PM |
PO09.00002: Analysis on the access condition to hybrid scenario in KSTAR Youngho Lee, SangKyeun Kim, Jaewook Kim, Minseo Park, Sang-hee Hahn, Boseong Kim, SeongMoo Yang, Cheol-Sik Byun, Woochang Lee, SeulChan Hong, Chan-Young Lee, Sangjin Park, Yong-Su Na After the successful operation of hybrid scenarios with a stationary performance of βN ≤ 3.0 and H89 ≤ 2.4 with the limited heating power Paux < 4.5MW [1], several experiments have been attempted to identify the robust access condition to the high-performance hybrid scenarios. In the 2020 campaign, it was found that the discharges with relatively high Dα baseline level and line averaged density could achieve the high performance after a performance bifurcation while other conditions were similar. The major cause is found to be the excitation of the edge localized coherent mode with mode number of n=6 and following broad-band fluctuations growing in the inter-ELM crashes. They enhance the transport in the edge region, which widens the pedestal and reduces its gradient, leading to the higher pedestal pressure [2]. Dedicated experiments to scan D2 fueling at the divertor region (using PVD) reveal that the increase of the Dα baseline level and edge density not only extied the edge localized coherent mode and broad-band fluctuations but also reproduce the high performance. |
Wednesday, November 10, 2021 2:24PM - 2:36PM |
PO09.00003: High li Steady State Scenario on KSTAR Jin Myung Park, Kyungjin Kim, Y.M. Jeon, H.S. Han, J.S. Kang, J. Chung, H.S. Kim, J. Ko, C.H. Holcomb, J.R. Ferron The high li scenario on KSTAR has achieved βN≈3, G=βNH89/q952≈0.3, Vloop≈0 at q95=5, close to the performance required for the ITER steady-state mission. This high li scenario relies on a very peaked current profile with qmin≈1 for the improved confinement and stability. The large positive magnetic shear from the peaked current profile leads to enhanced confinement and results in a high no-wall βN limit. The high βN condition can be achieved even without the need for wall stabilization. A long Ohmic or L-mode phase was used to allow current to penetrate to the axis, followed by injecting on-axis NB (4.3 MW) and EC (1.1 MW) power at the highest li≈1.5. A new robust high li recipe has been developed under the given superconducting coil environment of slow PF coil response to maintain shape control during the rapid βN ramp-up. High li≈1 discharges often develop small amplitude of n=2 MHD modes, resulting in the degradation of electron energy confinement. Good beam ion confinement was obtained with the estimated anomalous beam ion diffusion Db~0 both with and without the n=2 modes. The theory-based FASTRAN modeling with TGLF and EPED reproduces the experimental profiles reasonably well. Extrapolation to ITER will be discussed. |
Wednesday, November 10, 2021 2:36PM - 2:48PM |
PO09.00004: Control-Oriented Modeling via Transport Analysis for Advanced Scenario Control in KSTAR Zibo Wang, Eugenio Schuster, Tariq Rafiq, Hyun-seok Kim, Sang-hee Hahn, Jinil Chung, Jinseok Ko, Jisung Kang, Young-Mu Jeon, Will Wehner, David A Humphreys, Michael L Walker The regulation of the current profile, probably in combination with a measure of plasma internal energy, arises as one of the first fundamental control problems that must be solved to enable robust realization of advanced modes of operation in KSTAR. The high dimensionality and nonlinear magnetic-kinetic coupling of this problem motivate the use of a model-based control approach that can accommodate this complexity through embedding the known physics within the design. To achieve the scientific understanding of the current profile response dynamics in KSTAR that is necessary for control design, experiments have been recently carried out to characterize the response of the current profile to the different actuators. As an alternative to a purely data-driven modeling approach, response models suitable for model-based control design have been developed based on the Magnetic Diffusion Equation (MDE). This more physics-driven modeling approach is enabled by carrying out TRANSP-based transport analyses of the plasma-response data collected during the characterization experiments. The MDE is combined by either empirical scaling laws or transport equations for the electron density and temperature. Control applications enabled by this type of model, which include model-based optimal scenario planning, plasma state estimation from a limited set of noisy diagnostics, feedback control, and real-time prediction and optimization, will enhance present control capabilities in KSTAR. |
Wednesday, November 10, 2021 2:48PM - 3:00PM |
PO09.00005: Machine-Learning-based 0D Parameter Control on KSTAR Jaemin Seo, Yong-Su Na, Boseong Kim, Chanyoung Lee, Minseo Park, Sangjin Park, Youngho Lee It is required to control 0D plasma parameters such as βN, βp, q95, or li to sustain the fusion power and stability at the desired level in the future tokamak reactor. In this work, we describe a new scheme to control βN using the deep reinforcement learning (RL) technique that manipulates multi-dimensional control knobs such as plasma current and boundary shape parameters [1]. Deep RL algorithm trains an artificial decision-making agent to maximize the reward, through its trials and errors of virtual experiments in a data-driven simulator. As the reward is defined to increase when the achieved βN is close to the target, the RL agent is trained to provide the optimal scenario manipulation for the given target. We address several KSTAR experimental results conducted with RL-determined controls to achieve various targets of βN. They show successful control of βN by manipulating the plasma shape only, and also the achievement of high performance of βN~3 and H89~2.5. Furthermore, we trained the RL agent to control multiple 0D parameters, βp, q95, and li simultaneously into arbitrarily given target regimes, which can provide guidance of the operation scenario development [2]. This methodology can be a first step for the autonomous tokamak operation using machine learning techniques. |
Wednesday, November 10, 2021 3:00PM - 3:12PM |
PO09.00006: Tokamak Disruption Event Characterization and Forecasting Research and Expansion to Real-Time Application in KSTAR * Steven A Sabbagh, Young-Seok Park, John Berkery, James Bialek, Yanzheng Jiang, Veronika Klevarova, Juan D Riquezes, Jalal Butt, J.G. Bak, H.S. Han, Jayhyun Kim, Jinseok Ko, J.H. Lee, J.W. Lee, K. D Lee, Siwoo Yoon, Christopher Ham, Andrew Kirk, Lucy Kogan, David Ryan, Mark D Boyer, Keith Erickson, Mario L Podesta, Zhirui Wang, Fred M Levinton, Matt Galante Disruption prediction and avoidance is critical for ITER and reactor-scale tokamaks to maintain steady plasma operation and to avoid device damage. Physics-based disruption event characterization and forecasting (DECAF) provides early disruption forecasts (on transport timescales) that can be used for disruption avoidance through means such as profile control. In KSTAR, significant new hardware and software for real-time data acquisition and analysis are being installed including magnetics, plasma velocity and Te profiles, 2D internal Te fluctuations, and magnetic pitch angle. Real-time data has been taken for the first four with excellent agreement with offline data. DECAF analysis results are shown for multiple tokamaks including KSTAR, MAST, NSTX, and AUG. An NTM locked mode forecaster in DECAF using a torque balance model has been developed for off-line and real-time use. A new ELM event module includes the ability to distinguish local and global MHD. Supporting research includes pre-programmed ECCD used to reduce triggerless 2/1 mode amplitudes by ~80% the triggered mode amplitudes by 30%. TRANSP predict-first analysis show plasmas at βN > 3.5 with 100% non-inductive current drive. Ideal and resistive stability analyses using kinetic equilibrium reconstructions with MSE show sensitivity to local q and low shear regions. *US DOE grants DE-SC0020415, DE-SC0018623. |
Wednesday, November 10, 2021 3:12PM - 3:24PM |
PO09.00007: Torque balance analysis of rotating MHD for disruption prediction and avoidance in KSTAR Juan D Riquezes, Steven A Sabbagh, Young-Seok Park, John W Berkery, Veronika Klevarova, Yanzheng Jiang, Jalal Butt, J.G. Bak, J. Ko, Si-Woo Yoon, H.S. Han, Jayhyun Kim, J.W. Lee, Mark D Boyer, Keith Erickson Reactor scale tokamak devices require a low disruptivity rate to be a viable fusion energy producing system. An important precursor to disruptions is the presence of rotating MHD events that are often neoclassical tearing modes (NTM). Through electromagnetic and fluid drag torques, NTMs with a saturated island width can cause slowing of the mode and plasma rotation and lock it to the wall reference frame. A balance of the driving torque from the NBI, drag from perpendicular viscous diffusion drag and electromagnetic forces on the mode, and its inertia is used to model the rotation dynamics. Threshold rotation frequencies are derived from this model below which the mode rotation is expected to lead to a locking, serving as a disruption forecaster. Mode identification is computed most accurately by Fourier analysis of a toroidal array of magnetic probes or using simpler approaches more generally amenable to real-time calculation. From the rotation, the torque components are then calculated based on conditions for the expected drag torque ratios at the mode onset, changes in frequency, and Mirnov signal amplitudes. This technique will be employed for offline and real-time analysis of KSTAR plasmas with potential to signal use of active control or disruption mitigation systems |
Wednesday, November 10, 2021 3:24PM - 3:36PM |
PO09.00008: Sensitivity of EMC3-EIRENE Divertor Heat Flux Calculations to Variations of Magnetic Geometry from Plasma Response Calculations (GPEC) for applied RMP in KSTAR Jonathan M Van Blarcum, Heinke G Frerichs, Oliver Schmitz, Jong-Kyu Park, SeongMoo Yang, Tyler B Cote Good experimental agreement of a predictive model for Edge Localized Mode (ELM) suppression using Resonant Magnetic Perturbations (RMP) was observed at the Korean Superconducting Tokamak for Advanced Research (KSTAR) [J.-K. Park et al., Nature Physics 14 (2018)]. In this work the predicted ELM suppression window is computationally explored using a Field Line Analysis and Reconstruction Environment (FLARE) to calculate changes in the magnetic geometry and EMC3-EIRENE to calculate changes in the non-axisymmetric plasma edge solution and divertor heat flux. In the analysis, all plasma conditions are fixed except the magnetic field geometry which is a function of the RMP configuration and the subsequent plasma response, calculated by the Generalized Perturbation Equilibrium Code (GPEC). The calculated divertor heat flux is found to significantly change with the RMP definitions throughout the ELM suppression window, potentially advantageous for optimizing the divertor heat flux on ITER. However, the heat flux calculation was also found to be highly sensitive to the choice of the last rational surface in GPEC highlighting the importance of resolving a physically representative plasma response calculation for the purpose of making accurate heat flux predictions using EMC3-EIRENE. |
Wednesday, November 10, 2021 3:36PM - 3:48PM |
PO09.00009: Comparison of Pedestal Turbulence Dynamics in ELMing and RMP-Driven ELM Suppression Jaehyun Lee, Minho Kim, Gunsu Yun, Jae-Min Kwon, Juhyung Kim, Sumin Yi, Sehoon Ko, Minwoo Kim, Yongkyoon In With high-speed broadband electron cyclotron emission (ECE) measurements in KSTAR, the electron temperature turbulent fluctuations have been measured in high spatial resolution during the ELMing and ELM suppression phase. To investigate the structure and dynamics of turbulence occurring in the two different phases, spatial scales, parity, and cross-phase between electron temperature and radial velocity fluctuations were evaluated using velocimetry of measured 2-D electron temperature fluctuations. Furthermore, the effect of electron temperature fluctuations on maintaining the final form of the electron transport barrier was studied through comparative analysis with the profile data. Comprehensive comparisons with turbulence properties confirmed that the types and causes of turbulence in the two phases were different and that the role of contributing to heat and particle transport could be different as well. |
Wednesday, November 10, 2021 3:48PM - 4:00PM |
PO09.00010: Feedback Adaptive ELM Control towards (ITER) long pulse ELM suppression and its application to KSTAR Ricardo Shousha, SangKyeun Kim, Keith Erickson, Sang-hee Hahn, Oak Nelson, Josiah T Wai, SeongMoo Yang, Jong-Kyu Park, Egemen Kolemen Long pulse H-mode plasmas are typically conducive to the destabilization of Edge Localized Modes (ELMs), which can harm the fusion device beyond acceptable limits. Controlling these instabilities could prove crucial enabling economic fusion power. The objectives are to reduce the heat flux to the PFCs by suppressing ELMs through the application of 3D magnetic perturbations, while minimizing any performance degradation including noninductive fraction. To address this problem, a Feedback Adaptive ELM controller has been designed, implemented and tested on KSTAR. During the ELM suppressed phase, the controller adjusts 3D fields to reduce associated deleterious effects. Interestingly, we have found that edge collisionality could play an important role in such feedback loops. In addition, a precursor detector has been developed which informs the controller if a loss of ELM suppression is imminent. The controller can pre-emptively adjust the 3D perturbations in an attempt to minimize the loss of ELM suppression. Using this controller, we have been able to achieve ELM-suppression while maximizing non-inductive fraction, which is vital for long-pulse. Recently, the controller has been demonstrated on long pulses at KSTAR, showing promising results. |
Wednesday, November 10, 2021 4:00PM - 4:12PM |
PO09.00011: Characteristics of edge fluctuation and transport in the edge localized mode suppressed plasmas by the resonant magnetic perturbation field Minjun J Choi, Jae-Min Kwon, Juhyung Kim, Tongnyeol Rhee, Jun-Gyo Bak, Giwook Shin, Hyun-seok Kim, Kimin Kim, Byoung-Ho Park, Hogun Jhang, Gunsu S Yun, Minwoo Kim, Jong-Kyu Park, SangKyeun Kim, Hyung H Lee, Yongkyoon In, Jaehyun Lee, Min H Kim, Hyeon K Park The stochastic layer at the pedestal top by the resonant magnetic perturbation (RMP) field penetration has been known as a key mechanism in the RMP edge localized mode (ELM) suppression in tokamak plasmas. The local pressure profile flattening and broadband fluctuation increase characterize the pedestal top during the RMP ELM suppression. Despite the importance of turbulence to understand the pedestal transport, the origin of the increased fluctuation and its characteristics were not fully understood. Here, we suggest that the partially stochastic island at the pedestal top can play a significant role. The nonlinear mode coupling between an island and fluctuations can result in the fluctuation growth, and the stochastic fields around the island can make the spatial structure of turbulence less predictable. We used the bicoherence and the rescaled Jensen-Shannon complexity to analyze the three-wave coupling and the statistical characteristics of fluctuations, respectively. We show that (1) the weak but significant nonlinear mode coupling exists in the narrow region at the pedestal top and (2) the spatial structure of turbulence becomes less predictable around the pedestal top. These observations are not inconsistent with the expectation of the partially stochastic island. The latter can be used to estimate the effective width of the stochastic layer over which the characteristics of fluctuations and so transport are statistically distinguished from those over the normal field geometry. In addition, we found that the divertor particle flux near the striking point becomes less predictable with the more penetration of the RMP field into the plasma. |
Wednesday, November 10, 2021 4:12PM - 4:24PM |
PO09.00012: Machine learning-based preemptive RMP control for ELM-crash suppression in KSTAR plasmas Giwook Shin, Hyunsun Han, Minwoo Kim, Sang-hee Hahn, Wonha Ko, Gunyoung Park, Myungwon Lee, Youngho Lee, Minho Kim, Jae-wook Kim, June-woo Juhn, Ju-hyueok Jang, Hyun-seok Kim, Jongha Lee, Hajin Kim, Siwoo Yoon The ELM-crash suppression is a critical issue to ITER in that even a single ELM crash can severely reduce the lifetime of the plasma-facing components. The L-H transition and the onset of the ELMing occur within a somewhat predictable range but at an arbitrary moment that is difficult to know precisely. Due to such characteristics of the H-mode, the ELM-less H-mode can be achieved by automation of preemptive and event-driven control rather than feedforward approaches. Our innovative way [1] for the preemptive control is to apply RMP in the pedestal build-up period at low βN using a real-time machine learning algorithm. As a result, we can avoid the screening of RMP from the strong ExB and electron diamagnetic flow as expected from a fully developed H-mode pedestal at high βN. Furthermore, after applying RMP to the pedestal build-up period, the pedestal width is wider than the conventional RMP-ELM suppression discharges, while the gradient in the core region is relatively steep and high. Therefore, the shots sustain plasma performance steadily without rapid performance collapse or steady performance degradation except for the RMP effect. |
Wednesday, November 10, 2021 4:24PM - 4:36PM |
PO09.00013: RMP optimization on KSTAR SeongMoo Yang, Jong-Kyu Park, Nikolas C Logan, Y.M. Jeon, Qiming Hu, SangKyeun Kim, Yongkyoon In, W.H. Ko, Gunyoung Park, Jaehyun Lee, Jaewook Kim, Yong-Su Na A new systematic RMP optimization represents a step toward safe ELM suppression by isolating the edge from core resonant fields. The optimization identifies the edge-optimized RMPs (ERMPs) in principle from the external fields without core-resonant response but keeps its efficiency in ELM suppression with the actual coils for experimental implementation. The ERMP is validated with accurate predictions, where it provides a 100% increase of ELM suppression window compared to the standard RMP in KSTAR. Furthermore, the new scheme allows the optimization across scenarios such that it would suppress ELMs in the H-mode phase but remain stable in the L-mode phase. This integrated optimization enabled the application of 3D fields before the L-H transition and showed the possibility to control the ELMs through the entire periods of discharge, including the first ELM crash which alone could lead to significant damage to the plasma-facing components. It turns out that the ERMP penetrated the edge resonant surface already before L-H transition with the zero perpendicular flow and delayed L-H transition timing with decreased non-linear interaction between zonal flow and turbulence. This result highlights the need to integrate different resonant field physics through the entire discharge. |
Wednesday, November 10, 2021 4:36PM - 4:48PM |
PO09.00014: Non-Thermal "Cool" Fusion and Developments for the Ignitor Program Edoardo Boggio-Sella, Bruno Coppi, Gilberto Faelli, Piero Ferraris, Renato Spigler, Ignitor Program Members The Ignitor Program [1] has produced the first complete design of a machine capable of approaching ignition regimes with normally known conditions and acceptable safety factors. The ability of machines in this class, that include TRIAM-1 the first high field superconducting machine, to produce well confined plasmas with a wide range of collisionalities, is shown to be suitable at accessing alternative regimes including “cool fusion” where the excitation of radially “captive” ballooning modes can provide an efficient energy transfer from reaction products to fusing nuclei. The superconducting MgB2 technology that Ignitor has pioneered for the equilibrium coils remains adopted with recent advances. A collaboration with CNR laboratories on near term high field superconducting magnets is undertaken relating to relevant European institutions. The connection with outstanding departments of the Sapienza University involves materials study for other machine components, advances on sections of the present design and the fabrication of the central post, the central solenoid and the vertical field coils. |
Wednesday, November 10, 2021 4:48PM - 5:00PM |
PO09.00015: Fusion-Fission Concepts Based on Ignitor Technology and Physics Gilberto Faelli, Bruno Coppi, Cristina Mazzotta, Matteo Salvetti, Ignitor Program Members A development of the Ignitor program, aimed at making fusion reactors of near term interest, is that of starting from technology and physics advances emerged from the line of experiments that Ignitor represents to conceive novel hybrid reactors. This line has produced record high density plasmas with excellent confinement properties that can be utilized as D-T neutron sources, as suggested by E. P. Velikhov (2019), for power producing reactors with Thorium as its fissile component. The suggestion by C. Bolton (2020) to consider very high density pure-D, given that ignition conditions are not required, is taken into account as well. The Columbus concept [1] that had been studied as a follow-up to Ignitor in order to investigate the burn conditions of Tritium deprived plasmas is adopted as a reference device to start with. On this basis, a comprehensive analysis has been initiated in order to identify an optimal set of parameters for a D-T neutron source. Relevant advances in material science and fission reactor engineering are important factors involved in this analysis. |
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