Evaluation of layer potentials close to the boundary for Laplace and Helmholtz problems on analytic planar domains

DOI10.1137/120900253MaRDI QIDQ2874990zbMATH OpenOpenAlexFDO

Publication date 13 August 2014

Published in SIAM Journal on Scientific Computing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1310.5352, http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.251.3074

zbMATH Keywords

boundary integral equation convergence Helmholtz equation Laplace equation layer potential numerical example

Mathematics Subject Classification ID

Laplace operator, Helmholtz equation (reduced wave equation), Poisson equation (35J05) Stability and convergence of numerical methods for boundary value problems involving PDEs (65N12) Boundary element methods for boundary value problems involving PDEs (65N38)

Abstract: Boundary integral equations are an efficient and accurate tool for the numerical solution of elliptic boundary value problems. The solution is expressed as a layer potential; however, the error in its evaluation grows large near the boundary if a fixed quadrature rule is used. Firstly, we analyze this error for Laplace's equation with analytic density and the global periodic trapezoid rule, and find an intimate connection to the complexification of the boundary parametrization. Our main result is then a simple and efficient scheme for accurate evaluation up to the boundary for single- and double-layer potentials for the Laplace and Helmholtz equations, using surrogate local expansions about centers placed near the boundary. The scheme---which also underlies the recent QBX Nystr"om quadrature---is asymptotically exponentially convergent (we prove this in the analytic Laplace case), requires no adaptivity, generalizes simply to three dimensions, and has O(N) complexity when executed via a locally-corrected fast multipole sum. We give an example of high-frequency scattering from an obstacle with perimeter 700 wavelengths long, evaluating the solution at

2 i m e s 1 0^{5}

points near the boundary with 11-digit accuracy in 30 seconds in MATLAB on a single CPU core.

Review text

This article considers boundary integral equations, since they are an efficient and accurate tool to solve numerically elliptic boundary value problems. Usually, the solution is expressed as a layer potential, but a problem arises if one uses a fixed quadrature rule. Precisely, the error of the layer potential evaluation grows large close to the boundary. First, the author analyzes this error for the Laplace equation with analytic density using a global periodic trapezoidal rule. He finds an intimate connection to the complexification of the boundary parametrization. Next, a simple and efficient scheme is given that evaluates accurately up to the boundary the single- and double layer potentials both for the Laplace and the Helmholtz equation. The author uses a surrogate expansion about centers that are placed near the boundary. It turns out that the scheme is asymptotically exponentially convergent. The author proves this for the analytic Laplace case. Additionally, the scheme does not require adaptivity, it easily generalizes to three dimensions, and it has \(\mathcal{O}(N)\) complexity when executed via a locally corrected fast multipole sum. Finally, an example of high-frequency scattering for an obstacle that has perimeter 700 wavelength long is presented. The evaluation of the solution at 20000 points close to the boundary with 11-digit accuracy is achieved in 30 seconds using MATLAB on a single CPU core.

Recommendations

Cited in

(61)

Describes a project that uses

Uses Software

This page was built for publication: Evaluation of layer potentials close to the boundary for Laplace and Helmholtz problems on analytic planar domains

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2874990)