Regularity of Minimizers of Shape Optimization Problems involving Perimeter

TL;DR

This paper establishes existence and regularity of optimal shapes in perimeter-based shape optimization problems involving functionals like Dirichlet energy and spectral eigenvalues, under broad conditions.

Contribution

It proves regularity and existence of optimal shapes for perimeter minimization problems with complex functionals, extending previous results to more general settings.

Findings

Existence of optimal shapes under broad conditions.

Regularity results for minimizers of shape optimization problems.

Applicability to both unbounded and bounded domains.

Abstract

We prove existence and regularity of optimal shapes for the problem$$\min\Big\{P(\Omega)+\mathcal{G}(\Omega):\ \Omega\subset D,\ |\Omega|=m\Big\},$$where $P$ denotes the perimeter, $|\cdot|$ is the volume, and the functional $\mathcal{G}$ is either one of the following:\textless{}ul\textgreater{}\textless{}li\textgreater{} the Dirichlet energy $E\_f$, with respect to a (possibly sign-changing) function $f\in L^p$;\textless{}/li\textgreater{}\textless{}li\textgreater{}a spectral functional of the form $F(\lambda\_{1},\dots,\lambda\_{k})$, where $\lambda\_k$ is the $k$th eigenvalue of the Dirichlet Laplacian and $F:\mathbb{R}^k\to\mathbb{R}$ is Lipschitz continuous and increasing in each variable.\textless{}/li\textgreater{}\textless{}/ul\textgreater{}The domain $D$ is the whole space $\mathbb{R}^d$ or a bounded domain. We also give general assumptions on the functional $\mathcal{G}$ so that the result remains valid.