2006NCSU Libraries Repository (North Carolina State University Libraries)Requires access

A Distributed Lagrange Multiplier Based Fictitious Domain Method for Maxwell's Equations

Vrushali A. Bokil, R. Glowinski

Open publisher page 1 citations

Abstract

We consider a time-dependent problem of scattering by an obstacle involving the solution of the two dimensional Maxwell's equations in the exterior of a domain with a perfectly conducting condition on the boundary of this domain. We propose a novel fictitious domain method based on a distributed Lagrange multiplier technique for the solution of this problem. Perfectly matched layers are constructed to model the unbounded problem. Comparisons are performed with the finite difference scheme, that demonstrate the advantages of our fictitious domain method over the staircase approximation of the finite difference method. We conclude that our distributed multiplier approach is a simple, effective and far more accurate alternative to the popular FDTD method for solving Maxwell's equations.

About this research paper

What this paper is about

We consider a time-dependent problem of scattering by an obstacle involving the solution of the two dimensional Maxwell's equations in the exterior of a domain with a perfectly conducting condition on the boundary of this domain. We propose a novel fictitious domain method based on a distributed Lagrange multiplier technique for the solution of this problem. Perfectly matched layers are constructed to model the unbounded problem. Comparisons are performed with the finite difference scheme, that demonstrate the advantages of our fictitious domain method over the staircase approximation of the finite difference method. We conclude that our distributed multiplier approach is a simple, effective and far more accurate alternative to the popular FDTD method for solving Maxwell's equations.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

We consider a time-dependent problem of scattering by an obstacle involving the solution of the two dimensional Maxwell's equations in the exterior of a domain with a perfectly conducting condition on the boundary of this domain. We propose a novel fictitious domain method based on a distributed Lagrange multiplier technique for the solution of this problem. Perfectly matched layers are constructed to model the unbounded problem. Comparisons are performed with the finite difference scheme, that demonstrate the advantages of our fictitious domain method over the staircase approximation of the finite difference method. We conclude that our distributed multiplier approach is a simple, effective and far more accurate alternative to the popular FDTD method for solving Maxwell's equations.

Key concepts: Lagrange multiplier, Maxwell's equations, Fictitious domain method, Mathematics, Finite-difference time-domain method, Mathematical analysis, Boundary value problem, Domain (mathematical analysis)

Related papers

Back to paper searchBrowse research topicsOriginal source
A Distributed Lagrange Multiplier Based Fictitious Domain Method for Maxwell's Equations — Research Paper | ScholarLens