Session JI2: Invited MF: Pedestal, Edge-Localized Modes

\textbf{Super H-mode: Optimizing the pedestal to greatly raise performance in future fusion reactors}

The Super H-mode (SH) [1,2], a regime with high pedestal pressure and stored energy, has been developed into a reliably reproducible scenario on DIII-D using 3D fields to limit core mode growth. Unlike prior studies that employed counter-current beam injection with a quiescent H-mode edge, recent experiments have employed co-current beam injection at full magnetic field ($B_{T}=2.1-2.2T)$ and high current ($I_{P}=1.4-2.0MA)$, where discharges experience edge localized modes (ELMs). The discharges evolve from a high-performance transient phase that stores up to threefold higher energy density than conventional standard H-mode plasmas, into a stationary phase with $\beta_{N}\approx 2.5-2.9$, $W_{MHD}\approx 1.8-2.2MJ$, $\tau _{e}\approx 0.15-0.22s$, $H_{98}\approx 1.3-1.6$ , labelled the near Super H (NSH) regime. The transition from the very high confinement, high $T_{i}/T_{e}$ SH state at early times, to the reduced, but still high confinement NSH phase at lower $T_{i}/T_{e}$ has been studied with focus on limiting ion temperature gradient (ITG) turbulent transport and MHD instabilities. The transition correlates with a decrease in rotation shear and inductance (reducing beta limit and critical ITG) and either occurs through larger ELMs or abruptly by the destabilization of an internal mode. This global MHD event can cause energy losses of up to 20 {\%} of the stored energy. Typical H-mode instabilities do not show increased severity: overall moderate ELM sizes dominate with values below 5 {\%} relative to the total stored energy and present neoclassical tearing modes (NTMs) do not lead to disruptions. Progress towards divertor compatibility was achieved by decreasing ELM energy densities by 30{\%} through neon injection with constant or improved pedestal performance. The Super H-mode regime eases the operation margins for ITER to obtain Q$=$10 and is attractive for compact fusion reactor devices and possibly JET D-T. [1] Solomon PRL 2014 [2] Snyder NF 2015   

Internal measurement of pedestal-localized broadband magnetic fluctuations in ELMy H-mode plasmas in DIII-D

In DIII-D ELMy H-mode plasmas, pedestal-localized broadband magnetic fluctuations have been directly observed internally, for the first time, using a new Faraday-effect polarimeter diagnostic to identify their role in pedestal transport. The broadband magnetic fluctuations have many characteristics indicative of micro-tearing-modes (MTM): (a) poloidal wave number $\sim $0.3/cm, frequencies ranging from f$=$100-500 kHz with peak at 250 kHz, and propagation in the electron diamagnetic direction in the plasma frame, as expected for unstable MTM from linear GYRO calculation at the pedestal; (b) radial magnetic field amplitude lower bound $\vert \delta $B$_{\mathrm{r}}\vert \sim $25 Gauss and $\vert \delta $B$_{\mathrm{r}}$/B$\vert \sim $0.12$\% $ (B$=$2 T is total magnetic field) over bandwidth 100-500 kHz, comparable to the saturated MTM amplitude predicted by non-linear theory ($\rho _{\mathrm{e}}$/L$_{\mathrm{Te}}>$0.1$\% $ in pedestal); (c) non-monotonic dependence of mode amplitude on collision frequency, peaking at $\nu_{\mathrm{ei}}$/f$\sim $0.4-2 ($\nu_{\mathrm{ei}}$ is pedestal top collision frequency), consistent with lowest order MTM theory; (d): poloidally asymmetric spatial distribution with minimum amplitude near mid-plane. Between ELMs, the broadband magnetic fluctuation amplitude correlates with saturation of the pedestal gradients of T$_{\mathrm{e}}$, n$_{\mathrm{e}}$ and p$_{\mathrm{e}}$, indicating a role in regulating the pedestal. Based on stochastic field theory, the measured $\vert \delta $B$_{\mathrm{r}}\vert $ can lead to experimentally-relevant electron thermal transport while mode growth has been observed to correlate with decreased pedestal pressure and global stored energy. The observations provide strong evidence that MTM exists in H-mode pedestal and play an important role in pedestal transport. These findings provide critical experimental input for model validation and development of predictive physics understanding of pedestal confinement.   

Enhanced Pedestal H-mode Regime on NSTX

The largest normalized energy confinement on NSTX (H$_{\mathrm{98y,2}}$ \textgreater 1.5) was achieved in the ELM-free Enhanced Pedestal (EP) H-mode regime that features a wide pedestal with a significant increase in the edge T$_{\mathrm{i}}$ and rotation gradients. One feature of EP H-mode is a beneficial decrease in the impurity accumulation relative to standard ELM-free regimes. Recent analysis and comparison with 1-D transport models indicates that EP H-mode occurs when an increase in the anomalous pedestal transport reduces the edge density and collisionality such that the resulting improvement in the neoclassical ion thermal confinement exceeds the degradation driven by the larger anomalous transport. Linear CGYRO and GS2 calculations indicate the particle and electron energy transport is predominately due to TEMs in the steep gradient region and ETG modes contribute to the energy transport at the bottom of the pedestal. The ion energy transport is in good agreement with neoclassical transport throughout the pedestal. EP H-mode is often triggered by a period of reduced wall recycling following an ELM that leads to a temporary increase in the edge T$_{\mathrm{i}}$ and a corresponding reduction in the ion collisionality. The reduction in the neoclassical transport leads to an increase in the anomalous transport in all channels as the T$_{\mathrm{e}}$ profile is stiff, consistent with linear stability calculations and BES measurements. The increase in anomalous particle transport, combined with a reduction of the impurity pinch, reinforces the lower edge collisionality and can drive a positive feedback where the ion neoclassical energy confinement improvement exceeds the reduction due to anomalous transport. The enhanced thermal confinement at low pedestal ion collisionality motivates the development of actuators for controlling the edge density that are compatible with large core density and heat flux mitigation on NSTX-U.   

Poloidal variation and flow in I and H mode pedestals

A means of distinguishing between impurity accumulation in H mode and impurity removal in I mode operation based on poloidal flow, radial profiles, and poloidal variation measurements will be presented. The descriptions [1-3] are modifications of Helander's high Z impurity treatment [4]. They use poloidal impurity flow measurements rather than a main ion kinetic calculation of screening effects to determine the self-consistent poloidal variation of the impurity density and the electrostatic potential, and are thereby able to highlight the differing characteristics of I and H modes. In H mode the model predicts high (low) field side impurity accumulation when the poloidal flow and poloidal magnetic field are aligned (opposed) [1,2]. However, rotational effects must enter for I mode operation since $\mathrm{d}\ln T_z/\mathrm{d}\ln n_z$ is close to or greater than 2, where $n_z$ and $T_z$ are the impurity density and temperature, and the derivatives are radial [3]. The resulting outward impurity particle flux allows outboard impurity accumulation with the poloidal impurity flow in the direction opposite to the poloidal magnetic field close to the separatrix when $\mathbf{B} \times \boldsymbol{\nabla}B$ is away from the X-point. No weakly coherent mode need be present. The high Z treatment of Helander has also been extended to lower Z for large aspect ratio tokamaks by retaining the impurity diamagnetic effects that lead to impurity flows out of the flux surface. This extended large aspect ratio model allows consideration of the injected Boron behavior in H mode pedestals on Alcator C-Mod [5,6]. \begin{itemize} \item[[1]] S. Espinosa & P. J. Catto, Phys Plasmas 24, 055904 (2017) \vspace{-0.25cm} \item[[2]] S. Espinosa & P. J. Catto, PPCF 59, 105001 (2017) \vspace{-0.25cm} \item[[3]] S. Espinosa & P. J. Catto, PPCF 60, 094001 (2018) \vspace{-0.25cm} \item[[4]] P. Helander, Phys Plasmas 5, 3999 (1998) \vspace{-0.25cm} \item[[5]] C. Theiler et al., Nucl Fusion 54, 083017 (2014) \vspace{-0.25cm} \item[[6]] R. Churchill et al., Nucl Fusion 53, 122002 (2013) \end{itemize}   

The Dependence of the Impurity Transport on the Dominant Turbulent Regime in ELM-y H-mode discharges

Experiments and simulations on DIII-D in ELM-y H-mode plasmas demonstrate a large impact of a turbulence type on impurity particle transport, essential to reduce fuel dilution and avoid impurity accumulation in fusion devices. By changing the ratio of ion to electron heating using electron cyclotron (ECH) and neutral beam (NBI) heating in low collisionality plasmas, the electron to ion temperature ratio (T$_{\mathrm{e}}$/T$_{\mathrm{i}})$ is varied from 0.7 to 1.6, resulting in a clear transition from a turbulent regime dominated by Ion Temperature Gradient (ITG) mode to one with Trapped Electron Modes (TEM) and ITG mixture, as confirmed by gyrokinetic linear stability and changes in fluctuation measurements. Impurity transport for each T$_{\mathrm{e}}$/T$_{\mathrm{i\thinspace }}$value is probed by trace injections of Al and W ions using the recently installed laser blow-off system. Experimental transport coefficients are determined using the STRAHL code coupled with a Bayesian framework that is constrained by change exchange spectroscopy and soft-X ray measurements. The inferred diffusion of both impurity species increases by more than an order of magnitude outside of the ECH resonance compared to NBI only heated cases, and the change is clearly correlated with a transition from ITG to mixed ITG/TEM. Such a dramatic change in impurity flux is accompanied by mere 50{\%} increase of ion heat diffusion coefficients and a two-fold increase in electron heat diffusion. The reduced gyrofluid TGLF-SAT0 model quantitatively reproduces the increase in heat transport, but the change in impurity transport is underestimated across a large portion of the profile. In contrast, local non-linear gyrokinetic simulations performed with the CGYRO code match the heat flux as well as the impurity transport within experimental uncertainties.   

The Wavelet Nature of Persistent Edge Fluctuations Observed on Alcator C-Mod

Persistent edge fluctuations in high confinement regimes on Alcator C-Mod are found universally to be wavelet-like, exhibiting limited duration and toroidal extent. These findings, which emerge from analysis of a suite of poloidally and toroidally separated diagnostics, run counter to widely held assumptions that such fluctuations have a global character with long coherence length. Here, we examine the coherence time and length scales for three edge fluctuation phenomena observed on the C-Mod tokamak, existing in the range of $f\sim30-500$~kHz, $5\leq n \leq 30$, $k_{\perp} \rho_s<0.1$: the Quasi-Coherent Mode (QCM) of EDA H-mode, the Weakly Coherent Mode (WCM) of I-mode, and the inter-ELM Quasi-Coherent Fluctuation (QCF). Using data from Mirnov coils, phase contrast imaging, two-color interferometry, and reflectometry, we find that these modes have a relatively short coherence time (several periods), and similarly short coherence length (several wavelengths). The finite lifetime and toroidal extent of each coherent burst means that a disturbance may become strongly attenuated even before completing half a toroidal transit, consistent with the observation that fluctuation measurements made at the same poloidal, but opposing toroidal, angle are uncorrelated. The intermittent and localized nature of these bursts permits spatial asymmetries in the power spectra, and this can manifest itself in a surprising way: the QCM peak frequency is occasionally found to vary with toroidal angle for a portion of a discharge. These observations suggest the interpretation of these fluctuations as short, uncorrelated bursts of characteristic wavelets appearing at the edge, recalling WCM behavior observed on AUG by Manz \emph{et al.}, and point to the need for a unifying statistical language of intermittency to describe this form of edge turbulence, as applied to blob fluctuation spectra by Garcia \emph{et al.}