Laplacian eigenvalues for large negative Robin parameters on domains with outward peaks
Konstantin Pankrashkin, Firoj Sk, Marco Vogel
TL;DR
The paper analyzes the leading-order behavior of individual Robin Laplacian eigenvalues on domains with outward peaks as the Robin parameter becomes large. By reducing the problem to a cylindrical peak model and exploiting a transverse decomposition with the first cross-sectional Robin eigenfunction, the authors derive a sharp asymptotic expansion $\lambda_j(R^\Omega_\alpha)=A_\omega^{2/(2-q)}\lambda_j(L_1)\alpha^{2/(2-q)}+O(\alpha^{2/(2-q)-(q-1)})$, where $A_\omega=\mathcal{H}^{d-2}(\partial\omega)/\mathcal{H}^{d-1}(\omega)$ and $L_1$ is the one-dimensional operator $L_1 f= -f''+\big( {q^2(d-1)^2-2q(d-1) \over 4s^2}-1/s^q\big)f$. An Agmon-type estimate shows localization of eigenfunctions near the peak tip, and the analysis extends previous ball-cross-section results to general cross-sections $\omega$, illuminating how sharpness $q$ and geometry enter the asymptotics. The methods rely on variational principles, a cylindrical coordinate reduction, and careful truncation arguments to isolate peak contributions, with potential for generalizations to multiple peaks and non-isotropic settings.
Abstract
We study the asymptotic behavior of individual eigenvalues of the Laplacian in domains with outward peaks for large negative Robin parameters. A large class of cross-sections is allowed, and the resulting asymptotic expansions reflect both the sharpness of the peak and the geometric shape of its cross-section. The results are an extension of previous works dealing with peaks whose cross-sections are balls.
