MARS-F with the REORBIT module has been applied to study the runaway electron (RE) loss for a post-thermal quench JET discharge (95135@48.65s). The module has recently been improved to include the vacuum region and a surrounding limiter to allow more accurate modeling of RE loss. The results show that a fast growing kink-like n=1 mode located at r~0.85a and driven by plasma resistivity. Nearly full RE beam suppression was found with about kG field perturbation inside the plasma and within ~10ms time scale. RE loss was found to be generally not sensitive to the initial energy of REs, suggesting a relatively uniform loss of the REs at all energy levels. Spreading of lost REs on the JET limiter does not appear to be a monotonic function of field perturbation amplitude, implying that the loss pattern on the limiter surface depends on both the 3-D perturbation structure as well as amplitude.

Knowledge of accurate pellet ablation rates in plasmas is essential for modeling fueling and disruption mitigation in present and future tokamaks. In a collaboration with Stony Brook University, the 2D axisymmetric FronTier MHD code is used to simulate the influence of induction current forces on the ionized ablation cloud. In a previous paper, the code predicted that for a typical set of DIII-D plasma parameters the ablation rate of deuterium pellets was 22 g/s, considerably lower than the experimental values 39 g/s. Recently, considerable progress has been made to (i) introduce contact boundary conditions at the cloud/plasma interface, (ii) redesign the Riemann solvers, and (iii) refine the kinetic treatment for penetration of plasma electrons through the ablation cloud. The code is now getting 36 g/s, in good agreement with DIII-D experiments. In benchmarking with our 1D spherically symmetric transonic flow theory, the code found exact agreement for neon pellets over a broad range of pellet sizes and plasma conditions. A magnetic-field scan was carried out in 2D for neon pellets, fixing the electron temperature to 2 keV, the electron density to 10^14 cm^-3, and pellet radius to 2 mm. It was found that the ablation rate in 6-T fields is 6.6 times slower than the spherically symmetric theory prediction, and 2.6 times slower than it is in 2-T fields. This decrease in the ablation rate with magnetic field increases the feasibility of pellet fueling and disruption mitigation in high-field tokamak reactors. A thorough magnetic-field scan for deuterium will be reported shortly.

Two-fluid models incorporating energy or heat-fluxes have been implemented and tested in ALMA. The models, extensions of the famous Braginskii model, have not been widely utilized for plasma simulations before. The single-seeded blob propagation problem was used as a benchmark and shows qualitatively similar behavior to our previous simulations. The energy-flux model is rather challenging to simulate, among other reasons, because the highest order equation has an unknown closure. In this particular case, the closure was constructed as to replicate Braginskii’s diamagnetic heat-flux result. These models are currently being used as a testbed to evaluate more accurate reconstructions model closure using Gaussian Radial Basis Functions, with the final goal of incorporating self-consistent kinetic effects into a plasma fluid simulation.