The issues concerning the response of a plasma, at or near a singular
surface, to a magnetic perturbation with a phase velocity different fr
om the plasma flow velocity, are important for a number of phenomena.
Among these are ideal and nonideal magnetohydrodynamic stability of pl
asmas with shear flow or a flow relative to a resistive wall, sensitiv
ity of rotating plasma to field errors, and the ''locked mode'' phenom
enon. Models for the singular surface response have been tested agains
t results from ''magnetic braking'' experiments in DIII-D [R. J. La Ha
ye ct al., Nucl. Fusion 32, 2119 (1992)]. Previous models are found un
able to account for all of the experimental observations. A new heuris
tic nonlinear model presented in the paper may account for the observa
tions. A key element in the model is turbulence developed at the singu
lar surface; the turbulence is assumed driven by the singular layer di
ssipation and is assumed to impede the singular current through an ano
malous resistivity. When the perturbation amplitude is sufficiently la
rge, a positive feedback mechanism exists, since in the regime of inte
rest, dissipation increases with decreasing singular current. For smal
l perturbation amplitudes this mechanism is not operative so that prev
ious models for the response may be valid. (C) 1996 American Institute
of Physics.