arXiv subjects

R. Kube

Publications and source records attributed to R. Kube.

11 recordsLinked to original sources

arXiv2021

Machine learning accelerated particle-in-cell plasma simulations

Particle-In-Cell (PIC) methods are frequently used for kinetic, high-fidelity simulations of plasmas. Implicit formulations of PIC algorithms feature strong conservation properties, up to numerical round-off errors, and are not subject to time-step limitations which make them an attractive candidate to use in simulations fusion plasmas. Currently they remain prohibitively expensive for high-fidelity simulation of macroscopic plasmas. We investigate how amortized solvers can be incorporated with PIC methods for simulations of plasmas. Incorporated into the amortized solver, a neural network predicts a vector space that entails an approximate solution of the PIC system. The network uses only fluid moments and the electric field as input and its output is used to augment the vector space of an iterative linear solver. We find that this approach reduces the average number of required solver iterations by about 25% when simulating electron plasma oscillations. This novel approach may allow to accelerate implicit PIC simulations while retaining all conservation laws and may also be appropriate for multi-scale systems.

physics.plasm-ph↗

arXiv2020

Comparison between mirror Langmuir probe and gas puff imaging measurements of intermittent fluctuations in the Alcator C-Mod scrape-off layer

Statistical properties of the scrape-off layer (SOL) plasma fluctuations are studied in ohmically heated plasmas in the Alcator C-Mod tokamak. For the first time, plasma fluctuations as well as parameters that describe the fluctuations are compared across measurements from a mirror Langmuir probe (MLP) and from gas-puff imaging (GPI) that sample the same plasma discharge. This comparison is complemented by an analysis of line emission time-series data, synthesized from the MLP electron density and temperature measurements. The fluctuations observed by the MLP and GPI typically display relative fluctuation amplitudes of order unity together with positively skewed and flattened probability density functions. Such data time series are well described by an established stochastic framework which model the data as a superposition of uncorrelated, two-sided exponential pulses. The most important parameter of the process is the intermittency parameter, γ = τd / τw where τd denotes the duration time of a single pulse and τw gives the average waiting time between consecutive pulses. Here we show, using a new deconvolution method, that these parameters can be consistently estimated from different statistics of the data. We also show that the statistical properties of the data sampled by the MLP and GPI diagnostic are very similar. Finally, a comparison of the GPI signal to the synthetic line-emission time series suggests that the measured emission intensity can not be explained solely by a simplified model which neglects neutral particle dynamics.

physics.plasm-ph↗

arXiv2019

Outlier classification using Autoencoders: application for fluctuation driven flows in fusion plasmas

Understanding the statistics of fluctuation driven flows in the boundary layer of magnetically confined plasmas is desired to accurately model the lifetime of the vacuum vessel components. Mirror Langmuir probes (MLPs) are a novel diagnostic that uniquely allow to sample the plasma parameters on a time scale shorter than the characteristic time scale of their fluctuations. Sudden large-amplitude fluctuations in the plasma degrade the precision and accuracy of the plasma parameters reported by MLPs for cases in which the probe bias range is of insufficient amplitude. While some data samples can readily be classified as valid and invalid, we find that such a classification may be ambiguous for up to 40% of data sampled for the plasma parameters and bias voltages considered in this study. In this contribution we employ an autoencoder (AE) to learn a low-dimensional representation of valid data samples. By definition, the coordinates in this space are the features that mostly characterize valid data. Ambiguous data samples are classified in this space using standard classifiers for vectorial data. This way, we avoid to define complicate threshold rules to identify outliers, which requires strong assumptions and introduce biases in the analysis. Instead, these rules are learned from the data by statistical inference By removing the outliers that are identified in the latent low-dimensional space of the AE, we find that the average conductive and convective radial heat flux are between approximately 5 and 15% lower as when removing outliers identified by threshold values. For contributions to the radial heat flux due to triple correlations, the difference is up to 40%.

physics.plasm-ph↗

arXiv2018

Universality of Poisson-driven plasma fluctuations in the Alcator C-Mod scrape-off layer

Large-amplitude, intermittent fluctuations are ubiquitous in the boundary region of magnetically confined plasmas and lead to detrimental plasma-wall interactions in the next-generation, high duty cycle fusion power experiments. Using gas puff imaging data time series from the scrape-off layer in the Alcator C-Mod device, it is here demonstrated that the large-amplitude fluctuations can be described as a super-position of pulses with fixed shape and constant duration. By applying a new deconvolution algorithm on the data time series with a two-sided exponential pulse function, the arrival times and amplitudes of the pulses can be estimated and the measurement time series can be reconstructed with high accuracy. The pulse amplitudes are shown to follow an exponential distribution. The waiting times between pulses are uncorrelated, their distribution has an exponential tail, and the number of arrivals is a linear function of time. This demonstrates that pulse arrivals follow a homogeneous Poisson process. Identical statistical properties apply to both ohmic and high confinement mode plasmas, clearly demonstrating universality of the fluctuation statistics in the boundary region of Alcator C-Mod.

physics.plasm-ph↗

arXiv2018

Intermittent fluctuations in the Alcator C-Mod scrape-off layer for ohmic and high confinement mode plasmas

Plasma fluctuations in the scrape-off layer of the Alcator C-Mod tokamak in ohmic and high confinement modes have been analyzed using gas puff imaging data. In all cases investigated, the time series of emission from a single spatially-resolved view into the gas puff are dominated by large-amplitude bursts, attributed to blob-like filament structures moving radially outwards and poloidally. There is a remarkable similarity of the fluctuation statistics in ohmic plasmas and in edge localized mode-free and enhanced D-alpha high confinement mode plasmas. Conditionally averaged wave forms have a two-sided exponential shape with comparable temporal scales and asymmetry, while the burst amplitudes and the waiting times between them are exponentially distributed. The probability density functions and the frequency power spectral densities are self-similar for all these confinement modes. These results are strong evidence in support of a stochastic model describing the plasma fluctuations in the scrape-off layer as a super-position of uncorrelated exponential pulses. Predictions of this model are in excellent agreement with experimental measurements in both ohmic and high confinement mode plasmas. The stochastic model thus provides a valuable tool for predicting fluctuation-induced plasma-wall interactions in magnetically confined fusion plasmas.

physics.plasm-ph↗

arXiv2017

Relationship between frequency power spectra and intermittent, large-amplitude bursts in the Alcator C-Mod scrape-off layer

Fluctuations in the boundary region of the Alcator C-Mod tokamak have been analyzed using gas puff imaging data from a set of Ohmically heated plasma density scan experiments. It is found that the relative fluctuation amplitudes are modest and close to normally distributed at the separatrix but become increasingly larger and intermittent towards the main chamber wall. The frequency power spectra are nevertheless similar for all radial positions and line-averaged densities. Predictions of a stochastic model, describing the plasma fluctuations as a super-position of uncorrelated pulses, are shown to be in excellent agreement with the measurements. This implies that the pulse duration is the same, while the degree of pulse overlap decreases radially outwards in the scrape-off layer. The model also describes the rate of threshold level crossings, which provide novel predictions of plasma--wall interactions due to transient transport events.

physics.plasm-ph↗

arXiv2016

Amplitude and size scaling for interchange motions of plasma filaments

The interchange dynamics and velocity scaling of blob-like plasma filaments are investigated using a two-field reduced fluid model. For incompressible flows due to buoyancy the maximum velocity is proportional to the square root of the relative amplitude and the square root of its cross-field size. For compressible flows in a non-uniform magnetic field this square root scaling only holds for ratios of amplitudes to cross-field sizes above a certain threshold value. For small amplitudes and large sizes, the maximum velocity is proportional to the filament amplitude. The acceleration is proportional to the amplitude and independent of the cross-field size in all regimes. This is demonstrated by means of numerical simulations and explained by the energy integrals satisfied by the model.

physics.plasm-ph↗

arXiv2016

SOL width and intermittent fluctuations in KSTAR

Radial profiles of the ion saturation current and its fluctuation statistics are presented from probe measurements in L-mode, neutral beam heated plasmas at the outboard mid-plane region of KSTAR. The familiar two-layer structure, seen elsewhere in tokamak L-mode discharges, with a steep near-SOL profile and a broad far-SOL profile, is observed. The profile scale length in the far-SOL increases drastically with line-averaged density, thereby enhancing plasma interactions with the main chamber walls. Time series from the far-SOL region are characterised by large-amplitude bursts attributed to the radial motion of blob-like plasma filaments. Analysis of a data time series of several seconds duration under stationary plasma conditions reveals the statistical properties of these fluctuations, including the rate of level crossings and the average duration of periods spent above a given threshold level. This is shown to be in excellent agreement with predictions of a stochastic model, giving novel predictions of plasma-wall interactions due to transient transport events.

physics.plasm-ph↗

arXiv2015

Scrape-off layer turbulence in TCV: evidence in support of stochastic modelling

Intermittent fluctuations in the TCV scrape-off layer have been investigated by analysing long Langmuir probe data time series under stationary conditions, allowing calculation of fluctuation statistics with high accuracy. The ion saturation current signal is dominated by the frequent occurrence of large-amplitude bursts attributed to filament structures moving through the scrape-off layer. The average burst shape is well described by a double-exponential wave-form with constant duration, while the waiting times and peak amplitudes of the bursts both have an exponential distribution. Associated with bursts in the ion saturation current is a dipole shaped floating potential structure and radially outwards directed electric drift velocity and particle flux, with average peak values increasing with the saturation current burst amplitude. The floating potential fluctuations have a normal probability density function while the distributions for the ion saturation current and estimated radial velocity have exponential tails for large fluctuations. These findings are discussed in the light of prevailing theories for filament motion and a stochastic model for intermittent scrape-off layer plasma fluctuations.

physics.plasm-ph↗

arXiv2012

Burst statistics in Alcator C-Mod SOL turbulence

Bursty fluctuations in the scrape-off layer (SOL) of Alcator C-Mod have been analyzed using gas puff imaging data. This reveals many of the same fluctuation properties as Langmuir probe measurements, including normal distributed fluctuations in the near SOL region while the far SOL plasma is dominated by large amplitude bursts due to radial motion of blob-like structures. Conditional averaging reveals burst wave forms with a fast rise and slow decay and exponentially distributed waiting times. Based on this, a stochastic model of burst dynamics is constructed. The model predicts that fluctuation amplitudes should follow a Gamma distribution. This is shown to be a good description of the gas puff imaging data, validating this aspect of the model.

physics.plasm-ph↗

arXiv2007

Vector chiral order in frustrated spin chains

By means of a numerical analysis using a non-Abelian symmetry realization of the density matrix renormalization group, we study the behavior of vector chirality correlations in isotropic frustrated chains of spin S=1 and S=1/2, subject to a strong external magnetic field. It is shown that the field induces a phase with spontaneously broken chiral symmetry, in line with earlier theoretical predictions. We present results on the field dependence of the order parameter and the critical exponents.

cond-mat.str-el↗