Discovery of Density Limit Disruption Induced by Core-localized Alfv${é}$nic Ion Temperature Gradient Instabilities in a Tokamak Plasma

TL;DR

This work identifies core-localized Alfvénic ion temperature gradient (AITG) instabilities as a novel trigger for density-limit disruptions in HL-2A tokamak plasmas, appearing when the density nears the Greenwald bound ($n_e/n_{eG}>0.85$). Using a combination of core diagnostics and gyrokinetic stability analysis with GENE and the GFLDR framework, the study shows multi-band MHD activity localized between the $q=1$ and $q=2$ surfaces, consistent with AITG modes that destabilize and broaden outward, driving ion heat transport and triggering tearing modes that culminate in disruption. The results challenge edge-centric paradigms of density-limit disruptions and point to potential real-time AI-assisted disruption forecasting with tens of milliseconds lead times by monitoring core AITG activity. The findings imply a new, unified mechanism for density limits and suggest core-instability management as a pathway to safer, more predictable high-density operation in future reactors.

Abstract

To achieve a high energy gain, the fusion reactor plasma must reach a very high density. However, the tokamak plasmas ofen undergo disruption when the density exceeds the Greenwald density. The density limit disruption in tokamak plasmas is a mysterious barrier to magnetic confinement nuclear fusion, and hitherto, is still an unresolved issue. Over the past several years, the high density experiments with Greenwald density ratio $n_e/n_{eG}\sim1$ has been carried out using the conventional gas-puff fuelling method in HL-2A NBI and Ohmically heated plasmas. It is found for the first time that there are multiple-branch MHD instabilities in the core plasmas while $n_e/n_{eG}>0.85$. The simulation analysis suggests that the core-localized magnetohydrodynamics (MHD) activities belong to Alfv${é}$nic ion temperature gradient (AITG) modes, and on experiment firstly, it is discovered that they trigger the minor or major disruption of bulk plasmas while the density is peaked. These new findings are of great importance to figure out and understand the origin of density limit disruptions, as well as to forecast and avoid them for future fusion rectors.

Figures (11)

• Figure 1: A typical high-density disruption discharge with low-power NBI heating on HL-2A (ShotNo.38600). Plasma poloidal flux surface at $t=1565$$ms$, and the red dots show the locations of the core beam emission spectroscopy (Left col.). From top to bottom: plasma current ($I_p$), line averaged electron density ($ne$), standard gas puffing ($GP$), core electron temperature from TS ($Te$), NBI heating power ($P_{NBI}$), total radiation power ($P_{rad}$), and magnetic probe signals ($dB_{\theta}/dt$) (Right col.).
• Figure 2: Time evolution of the plasma density profile (a) and spectrogram of the core microwave interferometer signal (b)(ShotNo.38600). The LLM in the figure denotes the long-lived mode instability, with $m$ and $n$ being the poloidal and toroidal mode numbers of the instability, respectively. The observed instabilities during $t=1525-1578$$ms$ exhibit the distinct frequency dynamics, such as multiple bands, mode coexistence, frequency jumps and stair-like patterns.
• Figure 3: Time evolution of kinetic instabilities in high-density regime after NBI switched-off on HL-2A (ShotNo.38546). From top to bottom: plasma current ($I_p$), line averaged electron density ($ne$), NBI heating power ($P_{NBI}$), and spectrogram of the core microwave interferometer signal. The instabilities exhibit the distinct frequency dynamics, such as mode coexistence and structures resembling Christmas lights or mountain peaks.
• Figure 4: Plasma profiles (a-b), showing the electron temperature ($Te$), ion temperature($Ti$), toroidal rotation frequency ($f_{v\phi}$), electron density ($ne$), and safety factor ($q$), respectively (ShotNo.38600, t=1537 ms). Dependence of instability growth rates (c) and real frequencies (d) on wave-numbers ($k_y \rho_s$) at different radial positions ($\rho$), as given by GENE code.
• Figure 5: Dependence of real frequencies and growth rates of instabilities with different toroidal mode numbers on the safety factor $q$, as given by GFLDR based on experimental parameters where $\rho=0.3$.