NIMROD Shattered Pellet Injection (SPI) simulations involving scans over several parameters, including viscosity and toroidal deposition, found that, as expected, lower viscosity results in shorter thermal quench time due to a stronger linear response resulting in faster onset of (2,1) and (3,2) modes. SPI is the primary candidate for the disruption mitigation system on ITER. The simulations were aimed at studying the impact of various parameters on the quench dynamics. The final thermal collapse is usually dominated by a (1,1) core mode. These instabilities disrupt the plasma and are accompanied by a bright burst of radiation and rapid thermal collapse. Due to the instability driven thermal collapse, radiation efficiency for single injector is at best ~50%. In addition, it was found that, as the toroidal deposition width is decreased, the dynamics are delayed and the thermal quench occurs later. The higher concentration of a narrower toroidal deposition results in less overall ablation and assimilation. The peak radiation is also observed to decrease. For single injector SPI thermal quenches, the current spike is observed to occur several 100μs's after the plasma thermal quench. Symmetric multi-injector SPI simulations show that the current spike can be suppressed if the final dominant (1,1) core mode can be sufficiently suppressed. This is observed in NIMROD with simultaneous symmetric SPI injectors (e.g. placed 180º apart as in KSTAR's SPI system). Symmetric SPI injectors can suppress off-harmonic n-modes (e.g. tri-injector 120º apart, n/=3) and result in a more benign thermal quench (reduced radiation peaks and current spikes) with higher radiation efficiency ~80%.

Simulations of turbulence with the CGYRO gyrokinetic code have shown that at high density gradient the hydrogen energy flux can exceed the tritium energy flux due to finite electron to ion mass ratio effects from the strong nonadiabatic electron drive (E. Belli, J. Candy and R. E. Waltz, Phys. Plasmas 26 (2019) 082305). This is opposite to naive gyroBohm scaling and could contribute to the observed isotope scaling of energy confinement in tokamaks. The TGLF quasi-linear transport model was initially unable to match this result. Examination of the linear stability calculations in TGLF and CGYRO and the saturation model for the electric potential fluctuations in TGLF with Harry Dudding of the University of York, has identified the causes of the discrepancy. Changes to TGLF that increase collisional de-trapping and better model the effect of zonal flow mixing on the radial wavenumber spectral width bring TGLF into closer agreement with the linear and non-linear spectra of CGYRO. This enables TGLF to fully capture the key nonadiabatic electron dynamics to recover the isotopic dependence of the energy fluxes.