Deliberately applied 3D magnetic fields that tame the violent edge instabilities threatening fusion reactor walls
A resonant magnetic perturbation (RMP) is an intentionally applied, non-axisymmetric magnetic field whose helical pitch matches rational surfaces in the plasma edge. By breaking the perfect toroidal symmetry of the tokamak's confining field, RMPs interact with the plasma at resonant flux surfaces to modify the edge pressure gradient and current profile. Their primary application is the suppression or mitigation of edge-localized modes (ELMs)—violent, periodic bursts that expel energy from the plasma edge and can erode plasma-facing materials in reactor-scale devices.1
RMP coils, typically installed inside or just outside the vacuum vessel, generate magnetic field perturbations with toroidal mode numbers n = 1, 2, 3, or 4. The perturbation penetrates into the plasma edge and, depending on the plasma response, can either form magnetic islands at rational surfaces or induce shielding currents that partially screen the applied field. The net effect is to ergodize the magnetic field in the edge pedestal region, increasing radial transport and preventing the edge pressure gradient from reaching the peeling-ballooning stability boundary that triggers ELMs.2
RMP application is not without cost. The 3D field perturbation can cause density pump-out (a drop in pedestal density of 10–30%), reduce pedestal pressure, and drive toroidal rotation braking through neoclassical toroidal viscosity (NTV). These side effects reduce the energy confinement quality of the H-mode pedestal, partially offsetting the benefit of ELM control. Optimizing the RMP spectrum—choosing the right toroidal mode number, poloidal spectrum, and coil phasing—is an active area of research that depends sensitively on the q-profile, pedestal collisionality, and plasma rotation.4
The plasma response to RMPs is highly nonlinear. At low perturbation amplitude, the plasma screens the resonant field components; above a threshold, the field penetrates and island chains form. This bifurcation-like behavior makes predictive modeling challenging and is one reason why validated, multi-device scaling laws for RMP effectiveness remain an open research frontier in preparation for ITER operation.