Existence and regularity of optimal shapes for spectral functionals with Robin boundary conditions (Q2161494)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Existence and regularity of optimal shapes for spectral functionals with Robin boundary conditions
scientific article

    Statements

    Existence and regularity of optimal shapes for spectral functionals with Robin boundary conditions (English)
    0 references
    0 references
    4 August 2022
    0 references
    The author studies the problem of shape optimization, that is, minimization with respect to the (bounded, Lipschitz) domain \(\Omega \subset \mathbb{R}^n\), of functionals of the eigenvalues \(\lambda_k (\Omega;\beta)\) of the Laplacian with Robin boundary conditions in \(\Omega\), \[ \begin{cases} -\Delta u = \lambda u &\text{in } \Omega,\\ \partial_\nu u + \beta u =0 &\text{on } \partial\Omega, \end{cases} \] where \(\partial_\nu u\) is the outer normal derivative of \(u\), and \(\beta>0\) is a fixed positive constant. The paper under review contains two main results: the first states that for functionals \(F (\lambda_1, \ldots, \lambda_k)\) of the first \(k\) eigenvalues which are (strictly) increasing functions of each variable and which satisfy a certain coercivity condition in the \(k\)th variable (as well as certain other technical assumptions), the problem \[ \min \{ F(\lambda_1(\Omega;\beta), \dots, \lambda_k (\Omega;\beta) : \Omega \subset \mathbb{R}^n \text{ open }, |\Omega| \text{ fixed}\} \] always admits a solution in a suitable weak framework (the so-called SBV setting); moreover, any optimal set is rectifiable and its boundary has finite \(n-1\)-dimensional measure. A prototypical example of such a functional would be \(F(\lambda_1,\dots,\lambda_k) = \lambda_1 + \dots + \lambda_k\). Previously, the existence of minimizers was known only for the eigenvalues themselves, \(\lambda_k(\Omega;\beta)\); see [\textit{D. Bucur} and \textit{A. Giacomini}, J. Funct. Anal. 277, No. 3, 643--687 (2019; Zbl 1429.49015)]. The second major result concerns the minimizers \(\Omega\) of \(\lambda_k (\Omega;\beta)\) (for given \(k\)): the author shows that, in any dimension \(n \geq 3\), the \(k\)th eigenvalue of any minimizer \(\Omega\) in the given SBV framework is necessarily multiple: \[ \lambda_k (\Omega; \beta) = \lambda_{k-1} (\Omega;\beta). \] The corresponding problem for the more heavily studied and generally technically easier case of the Laplacian with Dirichlet boundary conditions has been open for around 20 years [\textit{A. Henrot}, Extremum problems for eigenvalues of elliptic operators. Basel: Birkhäuser (2006; Zbl 1109.35081)].
    0 references
    spectral optimization
    0 references
    Robin Laplacian
    0 references
    Robin boundary conditions
    0 references
    shape optimization
    0 references
    0 references
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references