Stellarators confine plasma using external magnetic coils shaped to produce a field with no net toroidal current. The advantage over tokamaks: no plasma current means no current-driven disruptions. The disadvantage: the field geometry must be precisely engineered to prevent particle drift losses. Quasi-axisymmetry does this by making the field strength look axially symmetric even though the geometry isn't — particles drift as if they're in a tokamak, staying confined.
But quasi-axisymmetric stellarators still develop bootstrap current — a self-generated current driven by pressure gradients and trapped particle dynamics. This current modifies the magnetic field, potentially destroying the carefully designed quasi-axisymmetry. Jorge et al. (arXiv: 2603.20125) propose a fix: break the stellarator into sections, each with a different omnigenous perturbation, designed so their bootstrap currents cancel.
The through-claim: piecewise symmetry can achieve what global symmetry cannot. A stellarator that is quasi-axisymmetric everywhere generates bootstrap current everywhere, and you're stuck with it. A stellarator that is quasi-axisymmetric in sections, with each section's perturbation chosen to generate equal and opposite bootstrap contributions, achieves zero net current while maintaining good confinement locally. The global optimization target (zero current) is met by local diversity (varied perturbations), not local uniformity.
This is a design principle: when a desired global property is incompatible with a single local structure, use multiple local structures whose contributions cancel. The solution isn't a better stellarator — it's a stellarator made of different stellarators.
Jorge, Landreman, Buller & Dorland, 2603.20125. Plasma physics / stellarators / omnigenity / bootstrap current.