Number of results to display per page
Search Results
72. Deep convolutional neural networks for multi-scale time-series classification and application to disruption prediction in fusion devices
- Author(s):
- Churchill, R.M; the DIII-D team
- Abstract:
- The multi-scale, mutli-physics nature of fusion plasmas makes predicting plasma events challenging. Recent advances in deep convolutional neural network architectures (CNN) utilizing dilated convolutions enable accurate predictions on sequences which have long-range, multi-scale characteristics, such as the time-series generated by diagnostic instruments observing fusion plasmas. Here we apply this neural network architecture to the popular problem of disruption prediction in fusion tokamaks, utilizing raw data from a single diagnostic, the Electron Cyclotron Emission imaging (ECEi) diagnostic from the DIII-D tokamak. ECEi measures a fundamental plasma quantity (electron temperature) with high temporal resolution over the entire plasma discharge, making it sensitive to a number of potential pre-disruptions markers with different temporal and spatial scales. Promising, initial disruption prediction results are obtained training a deep CNN with large receptive field ({$\sim$}30k), achieving an $F_1$-score of {$\sim$}91\% on individual time-slices using only the ECEi data.
- Type:
- Dataset
- Issue Date:
- October 2019
73. Enhanced Pedestal H-mode at low edge ion collisionality on NSTX
- Author(s):
- Battaglia, D.J.; Guttenfelder, W.; Bell, R.E.; Diallo, A.; Ferraro, N.;, Fredrickson, E.; Gerhardt, S.P.; Kaye, S.M.; Maingi, R.; Smith, D.R.
- Abstract:
- The Enhanced Pedestal (EP) H-mode regime is an attractive wide-pedestal ELM-free high-betap scenario for NSTX-U and next-step devices as it achieves enhanced energy confinement (H98y,2 > 1.5), large normalized pressure (betaN > 5) and significant bootstrap fraction (f_BS > 0.6) at I_p/B_T = 2 MA/T. This regime is realized when the edge ion collisionality becomes sufficiently small that a positive feedback interaction occurs between a reduction in the ion neoclassical energy transport and an increase in the particle transport from pressure-driven edge instabilities. EP H-mode was most often observed as a transition following a large ELM in conditions with low edge neutral recycling. It is hypothesized that the onset of pressure-driven instabilities prior to the full recovery of the neutral density leads to a temporary period with elevated ion temperature gradient that triggers the transition to EP H-mode. Linear CGYRO and M3D-C1 calculations are compared to beam emission spectroscopy (BES) and magnetic spectroscopy in order to describe the evolution of the edge particle transport mechanisms during the ELM recovery and the saturated EP H-mode state. The observations are consistent with the hypothesis that the onset of pressure-driven edge instabilities, such as the KBM and kink-peeling, can be responsible for the increased particle transport in EP H-mode.
- Type:
- Dataset
- Issue Date:
- June 2020
74. First impurity powder injection experiments in LHD
- Author(s):
- Nespoli F., Ashikawa N., Gilson E.P., Lunsford R., Masuzaki S., Shoji M., Oishi T., Suzuki C., Nagy A., Mollen A., Pablant N.A., Ida K., Yoshinuma M., Tamura N., Gates D.A., Morisaki T., and the LHD experiment group
- Abstract:
- Injection of impurities in the form of sub-millimeter powder grains is performed for the first time in the Large Helical Device (LHD) plasma, employing the Impurity Powder Dropper (IPD) [A. Nagy et al., RSI 2018], developed and built by PPPL. Controlled amounts of boron (B) and boron nitride (BN) powder are injected into the helical plasma. Visible camera imaging, UV and charge exchange spectroscopy measurements show that the injected impurities effectively penetrate into the plasma in two different magnetic configurations.The prompt effects of the impurities on the plasma are characterized as the injection rate is scanned. The injected impurities provide a supplemental electron source, causing the plasma density to increase, together with the radiated power. Beneficial effects on the confined plasma temperature are observed at low plasma densities, due to an increased efficiency in NBI power absorption. For $n_{e,av}<10^{19}m^{-3}$ the powder grains penetrate deeper into the plasma, as they can be less effectively deflected by the plasma flow in the divertor leg, which they have to cross first as they are injected from the top of the machine.In this case, the created B ions are observed to move outwards from UV spectroscopy and charge exchange measurements, due to the outwards direction of the radial electric field. This makes low density plasmas a better candidate for powder boronization techniques.
- Type:
- Dataset
- Issue Date:
- November 2020
75. Geometric concepts for stellarator permanent magnet arrays
- Author(s):
- Hammond, K. C.; Zhu, C.; Brown, T.; Corrigan, K.; Gates, D. A.; Sibilia, M.
- Abstract:
- The development of stellarators that use permanent magnet arrays to shape their confining magnetic fields has been a topic of recent interest, but the requirements for how such magnets must be shaped, manufactured, and assembled remain to be determined. To address these open questions, we have performed a study of geometric concepts for magnet arrays with the aid of the newly developed MAGPIE code. A proposed experiment similar to the National Compact Stellarator Experiment (NCSX) is used as a test case. Two classes of magnet geometry are explored: curved bricks that conform to a regular grid in cylindrical coordinates, and hexahedra that conform to the toroidal plasma geometry. In addition, we test constraints on the magnet polarization. While magnet configurations constrained to be polarized normally to a toroidal surface around the plasma are unable to meet the required magnetic field parameters when subject to physical limitations on the strength of present-day magnets, configurations with unconstrained polarizations are shown to satisfy the physics requirements for a targeted plasma.
- Type:
- Dataset
- Issue Date:
- July 2020
76. Gyrokinetic understanding of the edge pedestal transport driven by resonant magnetic perturbations in a realistic divertor geometry
- Author(s):
- Hager, R.; Chang, C. S.; Ferraro, N. M.; Nazikian R.
- Abstract:
- Self-consistent simulations of neoclassical and electrostatic turbulent transport in a DIII-D H-mode edge plasma under resonant magnetic perturbations (RMPs) have been performed using the global total-f gyrokinetic particle-in-cell code XGC, in order to study density-pump out and electron heat confinement.The RMP field is imported from the extended magneto-hydrodynamics (MHD) code M3D-C1, taking into account the linear two-fluid plasma response.With both neoclassical and turbulence physics considered together, the XGC simulation reproduces two key features of experimentally observed edge transport under RMPs: increased radial particle transport in the pedestal region that is sufficient to account for the experimental pump-out rate, and suppression of the electron heat flux in the steepest part of the edge pedestal.In the simulation, the density fluctuation amplitude of modes moving in the electron diamagnetic direction increases due to interaction with RMPs in the pedestal shoulder and outward, while the electron temperature fluctuation amplitude decreases.
- Type:
- Dataset
- Issue Date:
- June 2020
77. MHD-blob correlations in NSTX
- Author(s):
- Zweben SJ; Fredrickson ED; Myra JR; Podesta M; Scotti F
- Abstract:
- This paper describes a study of the cross-correlations between edge fluctuations as seen in the gas puff imaging (GPI) diagnostic and low frequency coherent magnetic fluctuations (MHD) in H-mode plasmas in NSTX. The main new result was that large blobs in the SOL were significantly correlated with MHD activity the 3-6 kHz range in 21 of the 223 shots examined. There were also many other shots in which fluctuations in the GPI signal level and its peak radius Rpeak were correlated with MHD activity, but without any significant correlation of the MHD with large blobs. The structure and motion of the MHD is compared with that of the correlated blobs, and some possible theoretical mechanisms for the MHD-blob correlation are discussed.
- Type:
- Dataset
- Issue Date:
- May 2020
78. Machine Learning Characterization of Alfvénic and Sub-Alfvénic Chirping and Correlation With Fast-Ion Loss at NSTX
- Author(s):
- Woods, B. J. Q.; Duarte, V. N.; Fredrickson, E. D.; Gorelenkov, N. N.; Podestà, M.; Vann, R. G. L.
- Abstract:
- Abrupt large events in the Alfvenic and sub-Alfvenic frequency bands in tokamaks are typically correlated with increased fast-ion loss. Here, machine learning is used to speed up the laborious process of characterizing the behavior of magnetic perturbations from corresponding frequency spectrograms that are typically identified by humans. The analysis allows for comparison between different mode character (such as quiescent, fixed frequency, and chirping, avalanching) and plasma parameters obtained from the TRANSP code, such as the ratio of the neutral beam injection (NBI) velocity and the Alfven velocity (v_inj./v_A), the q-profile, and the ratio of the neutral beam beta and the total plasma beta (beta_beam,i / beta). In agreement with the previous work by Fredrickson et al., we find a correlation between beta_beam,i and mode character. In addition, previously unknown correlations are found between moments of the spectrograms and mode character. Character transition from quiescent to nonquiescent behavior for magnetic fluctuations in the 50200-kHz frequency band is observed along the boundary v_phi ~ (1/4)(v_inj. - 3v_A), where v_phi is the rotation velocity.
- Type:
- Dataset
- Issue Date:
- December 2019
79. Modeling of resistive plasma response in toroidal geometry using an asymptotic matching approach
- Author(s):
- Z. R. Wang; A. H. Glasser; D. Brennan; Y. Q. Liu; J-K. Park
- Abstract:
- The method of solving linear resistive plasma response, based on the asymptotic matching approach, is developed for full toroidal tokamaks by upgrading the Resistive DCON code [A.H. Glasser, Z.R. Wang and J.-K. Park, Physics of Plasmas, \textbf{23}, 112506 (2016)]. The derived matching matrix, asymptotically matching the outer and inner regions, indicates that the applied three dimension (3-D) magnetic perturbations contribute additional small solutions at each resonant surface due to the toroidal coupling of poloidal modes. In contrast, the resonant harmonic only affects the corresponding resonant surface in the cylindrical plasma. Since the solution of ideal outer region is critical to the asymptotic matching and is challenging to be solved in the toroidal geometry due to the singular power series solution at the resonant surfaces, systematic verification of the outer region $\Delta^\prime$ matrix is made by reproducing the well known analytical $\Delta^{\prime}$ result in [H.P. Furth, P.H. Rutherford and H. Selberg, The Physics of Fluids, \textbf{16}, 1054-1063 (1073)] as well as by making a quantitative benchmark with the PEST3 code [A. Pletzer and R.L. Dewar, J. Plasma Physics, \textbf{45}, 427-451 (1991)]. Finally, the reconstructed numerical solution of resistive plasma response from the toroidal matching matrix is presented. Comparing with the ideal plasma response, the global structure of the response can be affected by the small finite island at the resonant surfaces.
- Type:
- Dataset
- Issue Date:
- October 2020
80. Simulating pitch angle scattering using an explicitly solvable energy-conserving algorithm
- Author(s):
- Zhang, Xin; Fu, Yichen; Qin, Hong
- Abstract:
- Particle distribution functions evolving under the Lorentz operator can be simulated with the Langevin equation for pitch angle scattering. This approach is frequently used in particle based Monte-Carlo simulations of plasma collisions, among others. However, most numerical treatments do not guarantee energy conservation, which may lead to unphysical artifacts such as numerical heating and spectra distortions. We present a novel structure-preserving numerical algorithm for the Langevin equation for pitch angle scattering. Similar to the well-known Boris algorithm, the proposed numerical scheme takes advantage of the structure-preserving properties of the Cayley transform when calculating the velocity-space rotations. The resulting algorithm is explicitly solvable, while preserving the norm of velocities down to machine precision. We demonstrate that the method has the same order of numerical convergence as the traditional stochastic Euler-Maruyama method.
- Type:
- Dataset
- Issue Date:
- September 2020
