Penalty combinations of the Ritz-Galerkin and finite difference methods for singularity problems (Q1362344)

From MaRDI portal





scientific article; zbMATH DE number 1043096
Language Label Description Also known as
default for all languages
No label defined
    English
    Penalty combinations of the Ritz-Galerkin and finite difference methods for singularity problems
    scientific article; zbMATH DE number 1043096

      Statements

      Penalty combinations of the Ritz-Galerkin and finite difference methods for singularity problems (English)
      0 references
      0 references
      20 April 1998
      0 references
      It is well known that incorporation of singular functions in the basis of the approximating subspace may greatly improve the accuracy of the finite element method [see results and references in the reviewer's book: Optimization in solving elliptic problems (1996; Zbl 0852.65087)]. In this paper, special attention is paid to a similar combination of difference methods (in \(S_1\)) and projective methods (in \(S_2\)), where \(\overline S_1\) and \(\overline S_2\) form a partition of the original region \(\overline S\) and \(\overline S_2\) contains a singular point. The coupling is done via a special penalty method on the cutting line \(\partial S_1\cap\partial S_2\). Estimates of type \(O(h^{3/2})\) are proved for a grid norm. Numerical experiments for a model problem are presented but the accuracy of approximations is not very high.
      0 references
      combined methods
      0 references
      error bounds
      0 references
      Ritz-Galerkin method
      0 references
      numerical experiments
      0 references
      singular functions
      0 references
      finite element method
      0 references
      difference methods
      0 references
      projective methods
      0 references
      penalty method
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references