The effectiveness of various realizations of the intelligent shell feedback stabilization concept in simulating an ideal conducting wall has been evaluated for DIII-D using the combined VACUUM and GATO codes. These realizations use perfect flux conserving coil arrays superimposed on the resistive vacuum vessel. It was found that one toroidally extended segment of coils alone with poloidal coverage around 12% of the vacuum tank surface area is effective in recovering up to 40% of the stabilizing effect of the ideal vessel. Increasing the number of segments to seven and the poloidal coverage up to 30% increases the effectiveness to 90%.

PIC simulations of ion rings in 3D using the FLAME code have revealed the vulnerability of such configurations to kink instabilities. This explains the typical helical ion beam structure observed in the FIREX experiments at Cornell. The instability results in an additional beam thermalization in the poloidal plane.

A simple improvement in the least square algorithm used in the TWIST-R code for extracting Δ’ from the resistive MHD eigenfunction resulted in a large increase in the accuracy of Δ’ near half rational values of the Mercier index μ (μ = 1/2, 1, 3/2, etc.) where the accuracy was previously poor. The new algorithm incorporates the symmetry in the large solution on the two sides of the rational surface so that the jump in the small solution, which is essentially Δ’, can be isolated accurately. Δ’ computed with the new algorithm now converges quadratically with mesh size, whereas, even though the eigenfunction converged quadratically, the original Δ’ values in TWIST-R converged only linearly.