“…Traditional numerical methods, such as the finite element method (FEM) [16,17,19,20], which is adaptive to the arbitrary-shaped problems, are now challenged by the problems such as excessive memory cost, long computation time, and limited modeling capability. However, based on the idea of Schwartz Method and Lagrange Multiplier, these methods such as FEM are now combined with a more efficient domain decomposition method (DDM) [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15]. Under this domain decomposition framework, large-scale EM problems can now be modeled with high parallel efficiency, especially with the aid of improved transmission conditions (TCs) at subdomain interfaces.…”
Abstract-In this paper, a novel second-order transmission condition is developed in the framework of non-conformal finite element domain decomposition method to meet the challenges brought by complex and large-scale electromagnetic modeling. First, it is implemented efficiently on the non-conformal interface via a Gauss integral scheme. Then, the eigenvalue analysis of the DDM system show a more clustered eigenvalue distribution of this transmission condition compared with several existing transmission conditions. After that, it is applied to large-scale complex problems such as S-type waveguides in the high frequency band and dielectric well-logging applications in the low frequency band. The final numerical results demonstrate that this transmission condition has high efficiency and huge capability for modeling large-scale problems with multi-resolution in any frequency band.
“…Traditional numerical methods, such as the finite element method (FEM) [16,17,19,20], which is adaptive to the arbitrary-shaped problems, are now challenged by the problems such as excessive memory cost, long computation time, and limited modeling capability. However, based on the idea of Schwartz Method and Lagrange Multiplier, these methods such as FEM are now combined with a more efficient domain decomposition method (DDM) [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15]. Under this domain decomposition framework, large-scale EM problems can now be modeled with high parallel efficiency, especially with the aid of improved transmission conditions (TCs) at subdomain interfaces.…”
Abstract-In this paper, a novel second-order transmission condition is developed in the framework of non-conformal finite element domain decomposition method to meet the challenges brought by complex and large-scale electromagnetic modeling. First, it is implemented efficiently on the non-conformal interface via a Gauss integral scheme. Then, the eigenvalue analysis of the DDM system show a more clustered eigenvalue distribution of this transmission condition compared with several existing transmission conditions. After that, it is applied to large-scale complex problems such as S-type waveguides in the high frequency band and dielectric well-logging applications in the low frequency band. The final numerical results demonstrate that this transmission condition has high efficiency and huge capability for modeling large-scale problems with multi-resolution in any frequency band.
“…Domain Decomposition Method (DDM) has been recognized as an important measure for designing efficiently computational algorithms [12][13][14][15][16].…”
Abstract-The hybrid finite element-boundary integral-multilevel fast multipole algorithm (FE-BI-MLFMA) is a powerful method for calculating scattering by inhomogeneous objects. However, the conventional FE-BI-MLFMA often suffers from iterative convergence problems. A non-overlapping domain decomposition method (DDM) is applied to FE-BI-MLFMA to speed up the iterative convergence. Furthermore, a preconditioner based on absorbing boundary condition and symmetric successive over relaxation (ABC-SSOR) is constructed to further accelerate convergence of the DDM-FE-BI-MLFMA. Numerical experiments demonstrate the efficiency of the proposed preconditioned DDM-FE-BI-MLFMA.
“…These methods in general require common boundaries between sub-domains and boundary conditions are enforced on sub-domain interfaces. There are usually two approaches used with the applications of the coupling effects: the direct method imposes the continuity of the fields on the partition interfaces and generates a global coupling matrix [17], whereas the iterative method [1,4] ensures the coupling between the adjacent elements by the transmission condition (TC) as described in [1]. It is possible to solve each sub-domain with the same method such as with finite element method (FEM) [4] or finite difference frequency domain method [8].…”
Section: Introductionmentioning
confidence: 99%
“…There are usually two approaches used with the applications of the coupling effects: the direct method imposes the continuity of the fields on the partition interfaces and generates a global coupling matrix [17], whereas the iterative method [1,4] ensures the coupling between the adjacent elements by the transmission condition (TC) as described in [1]. It is possible to solve each sub-domain with the same method such as with finite element method (FEM) [4] or finite difference frequency domain method [8]. However, some DDM methods have the flexibility that in each sub-domain the most efficient method can be used independently to solve Maxwell's equations [7].…”
Section: Introductionmentioning
confidence: 99%
“…A class of time and memory efficient algorithms is developed through which, the computational domain is divided into smaller sub-regions and then the sub-regions solutions, after introducing the effect of interactions between these sub-regions, are used to provide the entire domain solution. A group of methods that decomposes the computational domain into sub-domains is known as the domain decomposition methods (DDM) [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17]. These methods in general require common boundaries between sub-domains and boundary conditions are enforced on sub-domain interfaces.…”
Abstract-This paper presents a hybrid technique, which combines the desirable features of two different numerical methods, finite difference frequency domain (FDFD) and the method of moments (MoM), to analyze large-scale electromagnetic problems. This is done by dividing the computational domain into smaller sub-regions and solving each sub-region using the appropriate numerical method. Once each sub-region is analyzed, independently, an iterative approach takes place to combine the sub-region solutions to obtain a solution for the complete domain. As a result, a considerable reduction in the computation time and required computer memory is achieved.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.