The coupled current-and-pressure-driven instability at the plasma edge that triggers the explosive energy crashes known as ELMs.
A peeling-ballooning mode is an intermediate-n MHD instability that arises at the steep-gradient edge (pedestal) of a tokamak plasma operating in H-mode. It couples two distinct drives: the peeling component, an external-kink-like instability driven by the edge current density (bootstrap current), and the ballooning component, driven by the pressure gradient on the bad-curvature side. Neither drive alone is sufficient to trigger the mode under typical pedestal conditions; it is their coupling that defines the stability boundary and ultimately triggers edge-localized modes (ELMs).1
The peeling-ballooning framework was formalized in the EPED model (Edge Pedestal), which combines two constraints to predict the pedestal height and width before an ELM crash: (1) the kinetic-ballooning-mode (KBM) constraint that sets the pedestal width for a given height, and (2) the peeling-ballooning boundary that sets the maximum height for a given width. The intersection of these two constraints uniquely determines the pre-ELM pedestal structure.2
The stability boundary is conventionally plotted in a space of normalized pressure gradient (α) versus normalized edge current density (jedge). The peeling boundary runs roughly vertically at high current density; the ballooning boundary runs roughly horizontally at high pressure gradient. The unstable region occupies the upper-right corner. Plasma shaping—particularly triangularity, elongation, and squareness—expands the stable region, allowing higher pedestals and better confinement.3
Because the peeling-ballooning boundary determines when ELMs fire, all ELM-control strategies work by modifying the pedestal's trajectory through the stability diagram. Resonant magnetic perturbations (RMPs) ergodize the edge, limiting the pedestal gradient below the ballooning boundary. Pellet pacing triggers small, frequent ELMs before the pedestal reaches the natural crash point. Negative-triangularity and quasi-continuous-exhaust (QCE) regimes avoid the peeling-ballooning corner entirely by operating with different edge structures.4
The peeling-ballooning framework remains the workhorse theory for pedestal and ELM prediction in both present experiments and next-step reactor designs, though extensions to include diamagnetic, resistive, and kinetic effects continue to refine its quantitative accuracy.