A DRBEM approximation of the Steklov eigenvalue problem

In this study, we propose a novel approach based on the dual reciprocity boundary element method (DRBEM) to approximate the solutions of various Steklov eigenvalue problems. The method consists in weighting the governing differential equation with the fundamental solutions of the Laplace equation where the definition of interior nodes is not necessary for the solution on the boundary. DRBEM constitutes a promising tool to characterize such problems due to the fact that the boundary conditions on part or all of the boundary of the given flow domain depend on the spectral parameter. The matrices resulting from the discretization are partitioned in a novel way to relate the eigenfunction with its flux on the boundary where the spectral parameter resides. The discretization is carried out with the use of constant boundary elements resulting in a generalized eigenvalue problem of moderate size that can be solved at a smaller expense compared to full domain discretization alternatives. We systematically investigate the convergence of the method by several experiments including cases with selfadjoint and non-selfadjoint operators. We present numerical results which demonstrate that the proposed approach is able to efficiently approximate the solutions of various mixed Steklov eigenvalue problems defined on arbitrary domains.

Automated summary

This study proposes a novel approach based on the dual reciprocity boundary element method to approximate solutions for various Steklov eigenvalue problems. The method weights the governing differential equation with fundamental solutions of the Laplace equation, does not require interior nodes, and relates eigenfunctions to boundary flux in a novel way. It efficiently solves moderate-sized generalized eigenvalue problems at a smaller cost compared to full domain discretization alternatives.