Absorbing Boundary Conditions for Domain Decomposition

Abstract

There are striking similarities in the requirements for two classes of artificial boundary conditions. One consisting of those conditions used in computer simulations over unbounded domains and the other of those used in domain decomposition. In both cases the original problem does not contain these artificial boundary conditions and the differential equation is valid across the location of the artificial boundaries. The boundary condition should thus be as “transparent” as possible and not introduce significant errors. As we shall see below there are some differences between these classes of boundary conditions, but in this paper we shall stress the similarities and describe how techniques which were developed in connection to simulation over unbounded domains can be beneficial as boundary conditions in domain decomposition.KeywordsModel ProblemDomain DecompositionUnbounded DomainPseudo Differential OperatorDomain Decomposition MethodThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.