Book Description

The National Compact Stellarator Experiment (NCSX) is a three-field period compact stellarator presently in the construction phase at Princeton, NJ. The design parameters of the device are major radius R=1.4m, average minor radius a = 0.32m, 1.2 {le} toroidal field (B{sub t}) {le} 1.7 T, and auxiliary input power up to 12 MW with neutral beams and radio-frequency heating. The NCSX average aspect ratio R/a of 4.4 lies well below present stellarator experiments and designs, enabling the investigation of high {beta} physics in a compact stellarator geometry. Also the NCSX design choice for a quasi-axisymmetric configuration aims toward the achievement of tokamak-like transport. In this paper, we report on the magnetic field line tracing calculations used to evaluate conceptual plasma facing component (PFC) designs. In contrast to tokamaks, axisymmetric target plates are not required to intercept the majority of the heat flux in stellarators, owing to the nature of the 3-D magnetic field footprint. The divertor plate design investigated in this study covers approximately one half of the toroidal extent in each period. Typical Poincare plots in Figure 1 illustrate the plasma cross-section at several toroidal angles for a computed NCSX high-beta equilibrium. The plates used for these calculations are centered in each period about the elongated cross-section shown in Figure 1a, extending to +/- {pi}/6 in each direction. Two methods for tracing the edge field line topology were used in this study. The first entails use of the VMEC/MFBE-2001 packages, whereas the second entails use of the PIES code with a post-processor by Michael Drevlak; the same field line integration routine was used to evaluate the equilibria for this comparison. Both inputs were generated based on the {beta}=4%, =iota=0.5 equilibrium computed from the final NCSX coil set. We first compare these two methods for a specific plate geometry, and conclude with a comparison of the strike characteristics for two different target plate poloidal lengths using the latter method. The details of the magnetic topology differ when computed with VMEC/MFBE as compared with an iterated PIES solution. This difference is illustrated in Figure 2. The presence of islands in the PIES solution effectively reduces the radius of the last closed magnetic surface (LCMS) by about 8 cm. As expected, this difference in the edge topology translates to a difference in field line terminations.