Homogenization and nonselfadjoint spectral optimization for dissipative Maxwell eigenproblems
arXiv:2401.01049 · doi:10.1007/s00526-024-02908-0
Abstract
The homogenization of eigenvalues of non-Hermitian Maxwell operators is studied by the H-convergence method. It is assumed that the Maxwell systems are equipped with suitable m-dissipative boundary conditions, namely, with Leontovich or generalized impedance boundary conditions of the form . We show that, for a wide class of impedance operators , the nonzero spectrum of the corresponding Maxwell operator is discrete. To this end, a new continuous embedding theorem for domains of Maxwell operators is obtained. We prove the convergence of eigenvalues to an eigenvalue of a homogenized Maxwell operator under the assumption of the H-convergence of the material tensor-fields. This result is used then to prove the existence of optimizers for eigenvalue optimization problems and the existence of an eigenvalue-free region around zero. As applications, connections with the quantum optics problem of the design of high-Q resonators are discussed, and a new way of the quantification of the unique (and nonunique) continuation property is suggested.
31 pages, the discussion on connections with quantum optics and unique continuation is added, in particular, Example 2.2 (ii); 3 references are added; some of the proofs are shortened; typos are corrected
References in corpus (9)
- On uniqueness for time harmonic anisotropic Maxwell's equations with piecewise regular coefficients
- Long-lived Scattering Resonances and Bragg Structures
- From Steklov to Neumann via homogenisation
- Pareto optimal structures producing resonances of minimal decay under -type constraints
- Optimization of the lowest eigenvalue of a soft quantum ring
- M-dissipative boundary conditions and boundary tuples for Maxwell operators
- Pareto optimization of resonances and minimum-time control
- A Compactness Result for the div-curl System with Inhomogeneous Mixed Boundary Conditions for Bounded Lipschitz Domains and Some Applications
- Many Physical Design Problems are Sparse QCQPs