Session TO10: Magnetic Confinement: Edge & Pedestal Physics

Live

Using high-z wall materials attempts to switch the fusion challenge from heat load handling to removing impurities. We propose a means of measuring the radial impurity flux from currently available diagnostics, providing insight on optimal tokamak operation to prevent impurity accumulation [PoP 24, 055904 (2017)]. Our description is a modification of Per Helander's high Z impurity treatment [PoP 5, 3999 (1998)]. It uses poloidal impurity flow measurements rather than a main ion kinetic calculation of screening effects. High confinement mode operation was discovered 35 years ago to almost quadruple fusion power, and later explained by turbulence reduction by sheared flows. Less than a decade ago, improved mode operation was discovered to have the same desirable property, while removing impurities and providing fueling. Thanks to the impurity radial particle flux measuring technique developed, we explain the outward radial impurity flux without invoking a (sometimes undetected) turbulent mode [PPCF 60, 094001 (2018)]. This theory is supported by the observed ExB flow shear, which also explains the desired energy confinement via turbulence reduction.   

Live

Gyrokinetic (GK) models are increasingly being applied to the edge and scrape-off layer (SOL) of tokamaks in order to improve the description of parallel transport, trapped particles, and orbit loss effects that cannot be rigorously modeled in fluid models. Neutral interactions such as electron-impact ionization, charge exchange, and radiation play an important role in setting the plasma profiles, and it is necessary to include these to improve GK modelling of SOL plasmas. We describe a coupled continuum model for two gyrokinetic plasma species and a 5D (2X3V) monoatomic neutral species using the Gkeyll code. Our neutral model includes, so far, electron-impact ionization and charge exchange effects. The choice of a continuum model avoids the statistical noise and challenging convergence associated with hybrid continuum--Monte-Carlo codes. We present results from basic verification tests and describe progress towards realistic simulations of open-field-line turbulence in fusion devices.   

Live

The plasma transport in the open field line region beyond the separatrix, called the scape-off layer (SOL), is dominated by highly localized intermittent density filaments also called blobs. Originating from the interaction between magnetic curvature and E X B drifts, these blobs leak a sizable fraction of heat and particles perpendicular to the magnetic field towards the first-wall and can help spread out the power on a wider region of the divertor [1]. Using the Gkeyll computational plasma framework, we perform nonlinear electromagnetic simulations in helical, open field lines as a rough model for tokamak scrape-off layer [2], of the Axially Symmetric Divertor Experiment (ASDEX) Upgrade [3,4]. The simulation captures the spontaneous generation of the plasma blobs and produces significant particle and heat flux in the cross field direction towards the wall. In addition, the dynamically relaxed ion temperature profile attains a value higher than the electron temperature profile in the turbulent steady-state, quite similar to the experimental estimates. 1) D A D’Ippolito, et al, PoP, 18, 060501, 2011. 2) N R Mandell, et al, JPP, 86, 905860109, 2020. 3) B Nold, et al, PPCF, 52, 065005, 2010. 4) G Birkenmeier, et al, PPCF, 56, 075019, 2014   

Live

In a novel experiment on DIII-D, the behavior of inter-ELM pedestal instabilities responsible for particle and heat transport is probed by imposing fast vertical oscillations (jogs) on the entire plasma. These jogs induce current near the plasma edge, which is shown to couple with and modify the behavior of instabilities resident near the edge transport barrier. In the analyzed discharges, high-frequency modes appear after profile gradient clamping and increase in frequency as the plasma rotation velocity recovers. High-resolution magnetic and density fluctuation measurements localize these modes near the pedestal foot. After a jogging event, these microinstabilities appear at a higher frequency before slowing to their typical pre-ELM state while modes at the pedestal top remain unchanged. Analysis of charge-exchange data shows a significant overshoot in the post-jog recovery of the radial electric field well in the pedestal region that is associated with increased plasma rotation and mode frequency, showing correlation between the mode and the peak diamagnetic frequency. These observations point towards the classification of these modes as micro-tearing modes, in agreement with previous gyrokinetic work on DIII-D.   

Live

A novel high-speed spatial heterodyne spectrometer (SHS) for measuring internal plasma electric and magnetic fields has been installed on DIII-D. The diagnostic spectrally resolves the Balmer alpha neutral beam radiation split by the Motional Stark Effect (MSE). This provides information on the local magnetic field magnitude with the added benefit that the $E_r$ contamination and density sensitivity, present in polarimetry techniques, are significantly suppressed. This suppression and high optical throughput allow for very small changes in the edge magnetic field, and thereby local bootstrap current, to be measured. First measurements at rho~0.89-0.94 show a rapid increase in the local magnetic field ($\Delta B\approx{}$0.025T) during a period of high pressure ($\beta_N$ up to 2). Although inboard of the bootstrap current peak, these measurements indicate the potential sensitivity of the diagnostic to edge current changes. Forward modeling of the SHS spectrum using a core-MSE constrained kinetic EFIT is used to simulate spatial and temporal B-field changes induced by current density profile dynamics. In addition, significant changes of the local magnetic field strength are observed during ELM crashes.   

Live

The I-mode confinement regime is characterized by an H-mode like pressure pedestal and an L-mode like density edge profile, preventing impurity accumulation. A common feature of I-mode on Alcator C-Mod is a Low Frequency Edge Oscillation (LFEO) 8-30 kHz, concurrent with the ubiquitous Weakly Coherent Mode (100-300 kHz); the LFEO has been tentatively identified as the Geodesic Acoustic Mode by Gas Puff Imaging and a database analysis of central LFEO frequencies. High time resolution Electron Cyclotron Emission measurements have revealed two important spectral properties of the LFEO: a modulation of the frequency by the sawtooth cycle in which the frequency drops as the pedestal temperature rises, and an inverse dependence of the frequency on the impurity concentration of the plasma. The second observation is shown to be qualitatively consistent with two gyrokinetic GAM frequency scalings in the presence of impurity ions developed by Guo et al. and Sasaki et al. respectively, while the mechanism behind the first observation remains an open question. Supported by US DoE award SC0014264.   

Live

Developing improved reduced models of edge-pedestal plasma transport is an essential step towards better understanding, predicting, and optimizing tokamak performance. In the past, various drift-reduced fluid theories have been applied to model aspects relevant to the edge (e.g. pedestal formation, blob dynamics). As an initial step towards training against plasma diagnostics, this work explores the potential of validating and/or learning unobserved dynamics in 3-dimensional plasma turbulence directly from partial measurements. A synthetic plasma is numerically simulated using a reduced version of the two-fluid global drift-ballooning (GDB) code and carried out at edge reference densities, temperatures, and magnetic field relevant to boundary plasmas in high-field tokamaks using shearless field-aligned coordinates. Learning from partial observations is accomplished via physics-informed neural networks to accurately determine hidden dynamics of the radial electric field simply from measurements of turbulent electron density and temperature of limited spatial and temporal extent. This is a key stepping stone before training directly against experimental plasma discharges and demonstrates the ability to infer unmeasured quantities.   

Live

A comprehensive database of stationary DIII-D plasmas without ELMs compares all no-ELM regime types found in DIII-D: RMP-ELM suppression, QH-mode (incl. wide-ped), I-mode, EDA H-mode, regular L-mode, and negative triangularity L-mode (neg-D). Absolute plasma performance measured by Lawson product ($

Live

Recently, the negative triangularity tokamak (NTT) concept has received considerable attention because of several reasons: first, experimental results indicate that the plasma tends to stay in the low-confinement-mode (L-mode). If transitions into the high-confinement-mode (H-mode) are observed, the H-mode is characterized by high-frequency and small amplitude edge-localized-modes (ELMs). Furthermore, the energy confinement in an L-mode NTT has been observed to be close to or even above typical H-mode levels ($H_{98}(y,2) \geq 1$). These properties make the NTT an attractive option for a future fusion reactor.\\ In 2020, negative triangularity shapes including an active X-point have been developed on the ASDEX Upgrade tokamak. In order to avoid H-mode, these experiments were run with the ion grad B drift pointing away from the active X-point, and the value of the upper triangularity has been pushed to $\delta_{\rm upper} = -0.25$. Heating power scans have been performed encompassing different ion- to electron heating ratios. First results indicate good confinement and an L-mode typical edge with low impurity confinement and no large ELMs, in line with previous observations from other machines. An outlook indicating next steps and further research is given.   

Live

The pressure and temperature at the top of the pedestal play a key role in fusion performance of tokamaks. We employ an updated version of the EPED model to predict the pedestal structure, and derive a set of metrics to evaluate pedestal contributions to performance. We review comparisons of EPED predictions to observations on several tokamaks, focusing on high pedestal regimes such as Super H Mode, which has been explored in experiments on C-Mod and DIII-D [1-3]. EPED predictions, and DIII-D experiments in JET-similar shapes, suggest the possibility of Super H access on JET if sufficiently high triangularity shapes can be developed. The role of both plasma shaping and aspect ratio is studied in detail. Optimization of the pedestal presents opportunities to enable not only high fusion power density but also very high bootstrap current fraction, enabling compact devices with low recirculating power and continuous operation. A regime is identified with intermediate R/a=2.3-2.7, and strong shaping, which holds promise for next-generation fusion devices. [1] Hughes, J.W.., Nucl. Fus. 58 112003 (2018). [2] Snyder, P.B., Nucl. Fus. 59 086017 (2019). [3] Knolker, M., APS/DPP invited talk (2019), submitted to Phys. Plasmas.   

Live

We find that non-ideal effects can significantly affect peeling-ballooning stability thresholds in NSTX discharges. In particular, robust resistive peeling-ballooning modes are found to exist well before the ideal peeling-ballooning stability threshold is met. These results may help explain why ideal-MHD theory often does not accurately describe ELM onset in spherical torus configurations. The peeling-ballooning model is widely adapted to describe macroscopic stability limits related to the edge pedestal in tokamaks. The onset of such modes is a constraint in the EPED model, which is successfully applied to conventional aspect ratio machines to predict the pedestal height and width, where an onset of ELMs can be expected. For spherical tokamaks however the existing (ideal-MHD) models do not always accurately describe experimental observations. For the first time we present a study of peeling-ballooning modes beyond ideal-MHD based on experimental plasma configurations. Employing the state of the art extended-MHD code M3D-C1, we explore the impact of non-ideal effects on the linear stability of macroscopic modes in the plasma edge region of ELMing NSTX discharges. The results present a valuable basis for the development of a predictive model for ELMs in low-aspect ratio tokamaks.   

Live

Grassy ELM regime, a promising steady state operation regime for CFETR (China Fusion Engineering Test Reactor), has better confinement, robust impurity exhaust capability and broad heat flux width in EAST. Key parameters of the CFETR steady state scenario are in the range of experimental scaling law of Grassy ELM regime. However, the mechanism of the Grassy ELM remains unclear. To ensure the steady state operation of CFETR, we need to predict whether the steady state scenario is in the Grassy ELM regime. In this work, we use the BOUT$++$ code to simulate the onset of the ELM under CFETR steady state scenario. Linear simulation suggests the ballooning mode is unstable, and dominant toroidal mode number is 40. The ELM size is around 0.2{\%}, which is in the ELM size range of the Grassy ELM 0.1{\%}\textasciitilde 1{\%}. Comparing to the crash process of the Type-I ELM, initial crash, turbulence transport and saturation, our simulation results have a smaller initial crash, weaker magnetic perturbation and three phases of the turbulence transport stage. The three phases of the turbulence transport stage are dominant by multi-mode, high-n mode n$=$45 and low-n mode n$=$5, respectively. In the first two phase, the pressure perturbation peak inside the pedestal, while peak at the top of the pedestal at the last phase. To evaluate the erosion of the divertor target, the energy fluence at the outer divertor target is calculated, which is 0.029 MJ/m$^{\mathrm{2}}$, smaller than the tungsten melting limit 0.16 MJ/m$^{\mathrm{2}}$. .   

Live

Edge localized modes (ELMs) are routinely observed in H-mode plasma regimes of the National Spherical Torus Experiment (NSTX). Due to the rapid temporal evolution of this instability only diagnostics with high temporal and spatial resolution can provide a detailed insight into its dynamics. Gas-puff imaging (GPI) at NSTX can provide adequate 2D measurement of fast magnetic field aligned fluctuations (e.g. ELM filaments) in the scrape-off layer and at the plasma edge with 2.5us temporal and 1cm optical resolution. A novel analysis tool was developed to estimate the frame-by-frame velocities and structural parameters of the ELM filaments from GPI measurements. These methods were applied on 2010 NSTX H-mode discharges to characterize the dynamics of the ELM crash. The analysis revealed that the average ELM filament on NSTX has a peak radial velocity of 4km/s outwards and a peak poloidal velocity of 10km/s in the ion diamagnetic direction. The sizes of the ELM filaments were found to be similar to the filaments of the normal background turbulence. However, right at the ELM crash, their shape was found to be different, approximately circular. The ELM filament was found to onset 25us before the ELM crash time linearly accelerating radially along its path.   

Live

Comprehensive study, understanding and forecast of plasmas using direct numerical simulation requires an examination of scales vastly separated in space and time. Their analysis is facilitated by using models with varying levels of complexity. The Gkeyll framework aims to simulate plasmas at all scales with fluid, gyrokinetic and full kinetic models under the same infrastructure and interface. Gkeyll offers a multi-fluid moment model with a local collisionless closure and support for magnetosphere geometries, a gyrokinetic model for open-field line sheath-limited calculations, and a full kinetic Vlasov solver. Excluding the multi-fluid model, Gkeyll leverages and extends state-of-the art high-order discontinuous Galerkin methods to produce conservative and efficient algorithms. Gkeyll has been used for several physics investigations, like simulating Mercury's magnetosphere, analyzing electromagnetic effects in an NSTX-like scrape-off layer, and exploring magnetic field generation in 6D kinetic dynamos. We provide a description of Gkeyll and its capabilities, review recent physics results, and report on recent activities which include adding support for new geometries, neutral interactions, Vlasov-Poisson simulations, collisions with the Fokker-Planck operator, and using GPUs.