Robin eigenvalues on domains with peaks (Q2415286)

From MaRDI portal
Revision as of 06:53, 19 April 2024 by Importer (talk | contribs) (‎Changed an Item)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
scientific article
Language Label Description Also known as
English
Robin eigenvalues on domains with peaks
scientific article

    Statements

    Robin eigenvalues on domains with peaks (English)
    0 references
    0 references
    0 references
    21 May 2019
    0 references
    The authors consider the self-adjoint operator \( Q_{\alpha}\) in \(L^2(\Omega)\) associated with the quadratic form \[ \int_{\Omega}|\nabla u|^2dx-\alpha\int_{\partial\Omega}u^2d\sigma\,,\ \ \ u\in H^1(\Omega). \] Here \(\Omega\) is a domain in \(\mathbb R^N\) and \(\alpha>0\). With \(d\sigma\) is denoted the \(n-1\)-dimensional Hausdorff measure on \(\partial\Omega\). This operator is related to the Laplacian with Robin boundary conditions \(\partial_{\nu}u=\alpha u\) on \(\partial\Omega\), \(\nu\) denoting the outward unit normal. The authors are interested in the behavior of the bottom of the spectrum as \(\alpha\rightarrow+\infty\) in the case of a bounded Lipschitz domain with a single outward cusp of order less than quadratic. Namely, locally near the cusp the domain has the form \(\{(x_1,x')\in\mathbb R\times\mathbb R^{N-1}:x_1\in(0,\delta), |x'|< m x_1^p\}\) with \(p\in(1,2)\), \(m>0\). The requirement \(p<2\) ensures that \(E_1(Q_{\alpha})>-\infty\) for all \(\alpha>0\), where \(E_1(Q_{\alpha})\) is the first eigenvalue of \(Q_{\alpha}\). The requirement \(p>1\) ensures that \(\Omega\) has actually an outward cusp. They prove an asymptotic expansion of each single eigenvalue \(E_j(Q_{\alpha})\), namely \[ E_j(Q_{\alpha})=\left(\frac{\alpha}{m}\right)^{\frac{2}{2-p}}E_j(A_1)+o(\alpha^\frac{2}{2-p}), \] as \(\alpha\rightarrow+\infty\), where \(E_j(A_1)\) is the \(j\)-th eigenvalue of a suitable operator in \(L^2(0,\infty)\). The authors prove also an estimate of the number of eigenvalues below a moving threshold, namely \[ \mathcal N(Q_{\alpha},-B\alpha^{p+1})=M_p\alpha^{\frac{p-1}{2}}+o(\alpha^{\frac{p-1}{2}}) \] for some \(B>0\) as \(\alpha\rightarrow+\infty\), where \(M_p\) is explicit and depends only on \(N,p,m,B\).
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    asymptotic expansion
    0 references
    Robin boundary condition
    0 references
    Laplacian
    0 references
    negative eigenvalues
    0 references
    domains with peaks
    0 references
    0 references
    0 references