Three-dimensional modeling of plasma transport in the HIDRA stellarator

In the edge region of modern nuclear fusion experiments, the interactions between edge plasmas and the materials which ultimately confine them have become increasingly more important as device sizes and powers trend upwards. Devices such as MPEX and PISCES are built to investigate these interactions at the high energies and particle fluxes ejected by transient edge disruption events, but are ultimately linear devices. To extend the diagnostic environments of the greater fusion community to include a dedicated toroidal plasma-material interface (PMI) device, the Hybrid Illinois Device for Research and Applications (HIDRA) has been dedicated. HIDRA is an l=2, m=5 classical stellarator originally built in in the 1970's for RF heating studies. Its most recent users at the Max Planck Institute for Plasma Physics ran the device as WEGA to test heating schemes and train personnel for the recently-completed W-7X advanced stellarator, after which it was gifted to the University of Illinois at Urbana-Champaign. To improve the theoretical models of the PMI environment in the edge region of HIDRA, computational tools which apply these models are required. Many simulation tools currently in use focus on tokamak magnetic geometries or high-power, fully-ionized devices, necessitating the creation of an integrated suite of codes to handle partial ionization with more disparate operational power conditions and classical diffusivity unique to HIDRA in contemporary devices. To this end, HIDRAmod has been created, encompassing the existing coupled codes Edge Monte Carlo 3D (EMC3) and EIRENE to solve the plasma and neutral transport equations. FIELDLINES has been used in the creation of a field-aligned tetrahedral mesh generation utility TORMESH also integrated into HIDRAmod. In addition to these established codes, utilities for calculating the limiting surface and tetrahedral mesh intersection and for post-processing have been written. EMC3 has been altered to include a Bohm-like diffusivity to handle the uniquely diffusive plasma in a self-consistent manner. Preceding operational data on the device, simulations have been run under the context of bounding the incident heat and particle fluxes onto the limiting surface into a region of confidence based on parameters from previous operational campaigns. An outboard midplane limiter, inboard midplane limiter, and 'trench' limiter (along the bottom of the torus) were tested with RF input to core-edge power deposition efficiencies of 10-50% for a 26 kW 2.45 GHz combined RF input discharge. Axial magnetic field strengths of 87.5 mT and 0.5 T were analyzed, corresponding to two heating schemes tested at WEGA. Electron temperatures and densities were seen to match previous WEGA results of 8-10 eV and 1-3 · 10^12 cm^-3 in the edge region respectively. With these results, 26 kW of operational power translates to heat fluxes of up to 1 MW m^-2 on the inboard limiter, up to 0.2 MW m^-2 on the outboard limiter, and up to 0.15 MW m^-2 on the trench limiter. Particle fluxes have been similarly bound by upper limits of 4.7 · 10^22 m^-2 · s^-1, 5 · 10^21 m^-2 · s^-1, and 5.6 · 10^21 m^-2 · s^-1 for the inboard, outboard and trench limiters respectively. Scaling laws for peak electron temperature, Bohm-like diffusivity, and heat and particle fluxes have been calculated for both low- and high-field discharges; peak electron temperatures, particle diffusivity, and heat fluxes at the outboard limiter were seen to follow approximately a power-law of type f(P_RF) ∝ a · P_RF^b, with typical exponents in the range b ∼ 0.55 - 0.60. Higher magnetic fields have the tendency to linearize the heat flux dependence upon the RF power, with exponents in the range of b ∼ 0.75. Particle fluxes on the outboard limiter are seen to saturate first, and then slightly decline for RF powers in excess of 120 kW in the low-field case and 180 kW in the high-field case. Finally, extensions and applications of HIDRAmod are examined, including a path to a self-consistent full-device model and potential optimization strategies which may be employed to enhance fluxes arriving at the limiting surfaces