Table of Contents
Fetching ...

Anisotropic symmetrization, convex bodies, and isoperimetric inequalities

Gabriele Bianchi, Andrea Cianchi, Paolo Gronchi

TL;DR

The paper addresses a fully anisotropic variant of the Pólya–Szegő principle for functionals of the form $\int_{\mathbb{R}^n} \Phi(\nabla u) \,dx$, where $\Phi$ is an $n$-dimensional Young function and symmetry is taken with respect to a convex body $K$.It develops a direct, non-approximation proof based on anisotropic isoperimetric inequalities and Brunn–Minkowski theory, establishing the universal inequality $\int_{\mathbb{R}^n} \Phi_{\bullet K\bullet}(\nabla u^{K}) \,dx \le \int_{\mathbb{R}^n} \Phi(\nabla u) \,dx$ for $u \in V^{1,\Phi}_{\mathrm{d}}(\mathbb{R}^n)$ and equality characterization.Equality analysis shows that when equality occurs, $u$ is quasi-convex and its level sets and gradients satisfy precise geometric and variational relations involving $\Phi_{\bullet}$ and the symmetrized functionals, yielding necessary and (under mild convexity assumptions) sufficient conditions.By unifying isotropic and anisotropic Pólya–Szegő principles within fully anisotropic Orlicz–Sobolev spaces and providing a geometric, extremal-aware framework, the work offers new tools for anisotropic variational problems and sharp shape optimization results.

Abstract

This work is concerned with a Pólya-Szegö type inequality for anisotropic functionals of Sobolev functions. The relevant inequality entails a double-symmetrization involving both trial functions and functionals. A new approach that uncovers geometric aspects of the inequality is proposed. It relies upon anisotropic isoperimetric inequalities, fine properties of Sobolev functions, and results from the Brunn-Minkowski theory of convex bodies. Importantly, unlike previously available proofs, the one offered in this paper does not require approximation arguments and hence allows for a characterization of extremal functions.

Anisotropic symmetrization, convex bodies, and isoperimetric inequalities

TL;DR

The paper addresses a fully anisotropic variant of the Pólya–Szegő principle for functionals of the form $\int_{\mathbb{R}^n} \Phi(\nabla u) \,dx$, where $\Phi$ is an $n$-dimensional Young function and symmetry is taken with respect to a convex body $K$.It develops a direct, non-approximation proof based on anisotropic isoperimetric inequalities and Brunn–Minkowski theory, establishing the universal inequality $\int_{\mathbb{R}^n} \Phi_{\bullet K\bullet}(\nabla u^{K}) \,dx \le \int_{\mathbb{R}^n} \Phi(\nabla u) \,dx$ for $u \in V^{1,\Phi}_{\mathrm{d}}(\mathbb{R}^n)$ and equality characterization.Equality analysis shows that when equality occurs, $u$ is quasi-convex and its level sets and gradients satisfy precise geometric and variational relations involving $\Phi_{\bullet}$ and the symmetrized functionals, yielding necessary and (under mild convexity assumptions) sufficient conditions.By unifying isotropic and anisotropic Pólya–Szegő principles within fully anisotropic Orlicz–Sobolev spaces and providing a geometric, extremal-aware framework, the work offers new tools for anisotropic variational problems and sharp shape optimization results.

Abstract

This work is concerned with a Pólya-Szegö type inequality for anisotropic functionals of Sobolev functions. The relevant inequality entails a double-symmetrization involving both trial functions and functionals. A new approach that uncovers geometric aspects of the inequality is proposed. It relies upon anisotropic isoperimetric inequalities, fine properties of Sobolev functions, and results from the Brunn-Minkowski theory of convex bodies. Importantly, unlike previously available proofs, the one offered in this paper does not require approximation arguments and hence allows for a characterization of extremal functions.

Paper Structure

This paper contains 11 sections, 13 theorems, 199 equations.

Key Result

Theorem 3.1

Let $\Phi$ be an $n$-dimensional Young function. Assume that $u\in V^{1,\Phi}_{\rm d}({\mathbb R}^n)$. Then $u^{ K}\in V^{1,\Phi_{\bullet { K} \bullet}}_{\rm d}({\mathbb R}^n)$ and

Theorems & Definitions (29)

  • Theorem 3.1
  • Theorem 3.2
  • Example 3.3
  • Example 3.4
  • Theorem 3.5
  • Proposition 3.6
  • Proposition 3.7
  • Remark 3.8
  • Lemma 4.1
  • proof
  • ...and 19 more