Optimal thresholds for preserving embeddedness of elastic flows
Tatsuya Miura, Marius Müller, Fabian Rupp
TL;DR
The paper establishes sharp energy thresholds for preserving embeddedness of elastic flows of closed curves across codimensions. By formulating a variational problem for the scale-invariant energy $\bar{B}=LB$ on noninjective planar curves with unit rotation, the authors prove existence of a minimizer and derive a variational inequality whose analysis yields a complete classification of embedded cuspidal elasticae (ECEs) into a teardrop family and a heart-shaped family. In the planar case, the unique minimizer is the elastic two-teardrop, giving a precise threshold $C_{2T}$ that is strictly larger than the previously known $C_{8}$, while in higher codimensions the same threshold reduces to $C_{8}$, determined by the figure-eight elastica. The results translate to flow behavior: below the respective thresholds, elastic flows preserve embeddedness for all time and converge to circles (in the plane), and the thresholds are shown to be optimal by constructing embedded perturbations that develop self-intersections in finite time. The work thus links variational elasticae and geometric flow dynamics to yield exact, topology-aware energy bounds for embeddedness preservation.
Abstract
We consider elastic flows of closed curves in Euclidean space. We obtain optimal energy thresholds below which elastic flows preserve embeddedness of initial curves for all time. The obtained thresholds take different values between codimension one and higher. The main novelty lies in the case of codimension one, where we obtain the variational characterization that the thresholding shape is a minimizer of the bending energy (normalized by length) among all nonembedded planar closed curves of unit rotation number. It turns out that a minimizer is uniquely given by a nonclassical shape, which we call ``elastic two-teardrop''.
