Progress in understanding error-field physics in NSTX spherical torus plasmas

The low-aspect ratio, low magnetic field and wide range of plasma beta of NSTX plasmas provide new insight into the origins and effects of magnetic field errors. An extensive array of magnetic sensors has been used to analyse error fields, to measure error-field amplification and to detect resistive wall modes (RWMs) in real time. The measured normalized error-field threshold for the onset of locked modes shows a linear scaling with plasma density, a weak to inverse dependence on toroidal field and a positive scaling with magnetic shear. These results extrapolate to a favourable error-field threshold for ITER. For these low-beta locked-mode plasmas, perturbed equilibrium calculations find that the plasma response must be included to explain the empirically determined optimal correction of NSTX error fields. In high-beta NSTX plasmas exceeding the n = 1 no-wall stability limit where the RWM is stabilized by plasma rotation, active suppression of n = 1 amplified error fields and the correction of recently discovered intrinsic n = 3 error fields have led to sustained high rotation and record durations free of low-frequency core MHD activity. For sustained rotational stabilization of the n = 1 RWM, both the rotation threshold and the magnitude of the amplification are important. At fixed normalized dissipation, kinetic damping models predict rotation thresholds for RWM stabilization to scale nearly linearly with particle orbit frequency. Studies for NSTX find that orbit frequencies computed in general geometry can deviate significantly from those computed in the high-aspect ratio and circular plasma cross-section limit, and these differences can strongly influence the predicted RWM stability. The measured and predicted RWM stability is found to be very sensitive to the E × B rotation profile near the plasma edge, and the measured critical rotation for the RWM is approximately a factor of two higher than predicted by the MARS-F code using the semi-kinetic damping model.