The free-boundary curve shortening flow evolves a curve inside a convex domain, with endpoints free to slide along the boundary, shrinking the curve by its curvature at every point. Under this flow, a semi-circular arc in a convex domain contracts to a round half-point in finite time. The sharp convergence rate — meaning the estimate cannot be improved — characterizes exactly how fast the curve approaches its terminal shape as the singular time approaches.
Sharp rates in geometric flows are rare and structurally informative. A non-sharp estimate says the flow converges; a sharp estimate says there is no hidden slack in the convergence — the system approaches its limit as fast as geometry permits. The sharpness of the rate here means that the semi-circle is, in a precise sense, the most efficient path to extinction for this class of free-boundary evolutions. Any perturbation either converges at the same rate or slower.
Reader note (2026-11-17): Checked against the full text of arXiv:2603.06949 (Bourni, Burns and Langford; v1 of March 2026), which I had earlier held as a weak read. The headline matches: a sharp rate of convergence of the rescaled curves to the shrinking semicircle. Three details differ from the essay. The paper's setting is a maximal free-boundary flow starting from a properly embedded curve in a convex domain, not from a semicircular arc; it cites an earlier result that such a flow either converges in infinite time to a critical chord or shrinks in finite time to a round half-point, and its theorem concerns the finite-time case, where the rescaled curves converge to the unit semicircle in every C-k norm at a rate no slower than (T minus t) to the power 1 minus delta, for any delta greater than zero. The proof is a stability analysis: the linearization at the semicircle has unstable modes corresponding to time translation and horizontal translation, which the authors remove by fixing the enclosed area and a center of mass. And the paper remarks that in the special case where the domain is a hyperplane (the paper's wording) the rate improves to 2 minus delta. The essay's reading of sharpness as the semicircle being the most efficient path to extinction, and of the semicircle as the flow's chosen attractor, is the essay's interpretation of that result.
The general principle: in systems that evolve toward a singularity, the shape that achieves the singularity most efficiently reveals the geometry's intrinsic preference. The semi-circle is not just one possible extinction profile — it is the attractor, the shape that the flow's internal logic selects. Optimality under extinction exposes the flow's hidden symmetry.
(arXiv:2603.06949)