RUS  ENG
Full version
JOURNALS // Vestnik Yuzhno-Ural'skogo Universiteta. Seriya Matematicheskoe Modelirovanie i Programmirovanie // Archive

Vestnik YuUrGU. Ser. Mat. Model. Progr., 2012 Issue 14, Pages 130–140 (Mi vyuru89)

This article is cited in 2 papers

Mathematical Modelling

Construction of Quasi-Structured Locally Modified Grids for Solving Problems of High Current Electronics

V. M. Sveshnikov, D. O. Belyaev

Institute of Computational Mathematics and Mathematical Geophysics of the Siberian Branch of the Russian Academy of Sciences (Novosibirsk, Russian Federation)

Abstract: An algorithm for constructing quasi-structured grids, which consist of the uniform rectangular subgrids and are built in two stages, is considered. At first, the computational domain is covered by a uniform rectangular macrogrid, and then an original uniform rectangular subgrid is set for each macroelement. It is essential that subgrids can be inconsistent. By adjusting a density of the subgrid nodes an adaptation of the quasi-structured grid to inhomogeneities within the domain is achieved. To adapt the grid to the exterior boundary we make local modifications consisting in a shift of the near boundary nodes to the boundary. Here we propose the algorithm of the local modification to construct a quasi-structured grid of high quality which does not break a subgrid structuring. The suggested quasi-structured grids profitably differ from structured grids in that they do not require extra nodes to support structuring and, also, storing a large amount of information as for unstructured ones. The solution of boundary value problems for quasi-structured grids is found by a decomposition of the computational domain into subdomains without overlapping. This method can be easily parallelized and, therefore, used to carry out calculations on multiprocessor supercomputers.

Keywords: quasi-structured grids, local modification, domain decomposition method, boundary value problems, high-current electronics.

UDC: 519.63

MSC: 65N50

Received: 05.02.2012



© Steklov Math. Inst. of RAS, 2024