Divertor Detachment Characterization in Negative Triangularity Discharges in DIII-D via 2D Edge-Plasma Transport Modeling

TL;DR

This work investigates why negative triangularity divertor configurations on DIII-D have difficulty achieving deep divertor detachment. Using the 2D edge-plasma transport code UEDGE, including cross-field drifts, the authors model NT discharges with neutral-beam heating and compare them to Ohmic NT and PT cases, revealing that detachment onset occurs at higher upstream densities for NT due to shorter connection lengths, smaller divertor volume, and reduced cross-field transport. The simulations reproduce experimental trends: forward BT requires higher densities to detach than reverse BT, and reverse BT fails to reach deep detachment, instead undergoing a thermal collapse; PT detaches more readily at lower densities. The study also isolates geometry and transport contributions to detachment thresholds, finding that extended divertor legs and enhanced transport reduce NT-PT detachment density gaps, with the remaining difference likely from the intrinsically shorter NT connection length. Overall, the work provides a physics-based explanation for the detachment behavior in NT discharges and highlights the roles of divertor geometry, cross-field transport, and E$\times$B dynamics in determining detachment access.

Abstract

Edge fluid modeling of the first divertor-plasma detachment experiments in negative triangularity discharges on DIII-D is presented using the 2D multi-fluid code UEDGE, including cross-field particle drifts. Density scans are performed to reproduce the experimental roll-over of the outer-target ion saturation current and to investigate detachment physics for both forward and reverse toroidal magnetic field configurations. Consistent with experiments, the simulations show that approximately 40% higher density is required to reach detachment onset for forward BT compared to reverse BT, and that deep detachment is not achieved for reverse BT. Comparisons with positive triangularity Ohmic discharges further demonstrate that negative triangularity requires substantially higher densities, at or above the Greenwald limit, to access detachment. The modeling indicates that the increased difficulty of achieving detachment in negative triangularity arises from a shorter midplane-to-target connection length, a reduced outer divertor leg length, and lower cross-field transport compared to positive triangularity configurations.

Figures (15)

• Figure 1: UEDGE grid of the DIII-D negative triangularity discharge (#194288@3000ms) with neutral beam injection $P = 4.3~\mathrm{MW}$ and separatrix intersecting the vessel wall at the location of the armor tiles.
• Figure 2: Left: OSP $J_\mathrm{sat}$ as a function of run time for both the forward (red) and reverse (black) $B_T$ cases; Middle: Line-averaged electron density as a function of run time for both the forward (red) and reverse (black) $B_T$ cases; Right: OSP $T_e$ as a function of run time for both the forward (red) and reverse (black) $B_T$ cases. The vertical dashed lines denote the time at which $J_\mathrm{sat}$ reaches its maximum (typically associated with detachment onset) corresponding to line densities of $1.3\times n_G$ and $1.0\times n_G$ for the forward (red) and reverse (black) $B_T$ cases, respectively, where $n_G$ is the Greenwald density.
• Figure 3: Top row: Thomson scattering data from NT-campaign with NBI (dots) for (a), electron temperature, (b), electron density, and (c), $n_{C^{6+}}$ density, and UEDGE-computed profiles (solid) at the outer midplane for forward (red) and reverse (black) $B_T$. Second row: UEDGE input for (d), electron/ion thermal diffusivities, (e), deuterium particle diffusivity, and (f), carbon particle diffusivity and pinch velocity (dashed in m/s).
• Figure 4: Electron temperature (left) and density (right) along the outer target plate from both the experiment (dots) and UEDGE simulation (solid) for both the forward (red) and reverse (black) $B_T$ configurations
• Figure 5: Left: Integrated ion saturation currents ($I_\mathrm{sat}$) with the increase of outer midplane separatrix electron density for both the forward (red) and reverse (black) $B_T$ configurations from UEDGE simulations. The two vertical dashed lines denote $n_\mathrm{e,sep}$ at the roll-over of OSP $J_\mathrm{sat}$ observed in the experiments for the forward (red) and reverse (black) $B_T$ configurations. Right: Peak electron temperature on the outer target plate as the increase of outer midplane separatrix electron density for both the forward (red) and reverse (black) $B_T$ configurations from UEDGE.