Fuglede-type arguments for isoperimetric problems and applications to stability among convex shapes

Abstract

This paper is concerned with stability of the ball for a class of isoperimetric problems under convexity constraint. Considering the problem of minimizing $P+\varepsilon R$ among convex subsets of $\mathbb{R}^N$ of fixed volume, where $P$ is the perimeter functional, $R$ is a perturbative term and $\varepsilon>0$ is a small parameter, stability of the ball for this perturbed isoperimetric problem means that the ball is the unique (local, up to translation) minimizer for any $\varepsilon$ sufficiently small. We investigate independently two specific cases where $\Omega\mapsto R(\Omega)$ is an energy arising from PDE theory, namely the capacity and the first Dirichlet eigenvalue of a domain $\Omega\subset\mathbb{R}^N$. While in both cases stability fails among all shapes, in the first case we prove (non-sharp) stability of the ball among convex shapes, by building an appropriate competitor for the capacity of a perturbation of the ball. In the second case we prove sharp stability of the ball among convex shapes by providing the optimal range of $\varepsilon$ such that stability holds, relying on the \emph{selection principle} technique and a regularity theory under convexity constraint.