A preconditioner for the \(h\)-\(p\) version of the finite element method in two dimensions (Q2564476)

From MaRDI portal





scientific article; zbMATH DE number 966628
Language Label Description Also known as
default for all languages
No label defined
    English
    A preconditioner for the \(h\)-\(p\) version of the finite element method in two dimensions
    scientific article; zbMATH DE number 966628

      Statements

      A preconditioner for the \(h\)-\(p\) version of the finite element method in two dimensions (English)
      0 references
      0 references
      0 references
      16 October 1997
      0 references
      The paper describes a preconditioner for the \(h\)-\(p\) version of the finite element method in two dimensions. It assumes a two-level mesh: the coarse level mesh with triangular or quadrilateral elements (called also substructures \(\Omega_i\) with diameter \(H_i)\) equipped with linear or bilinear shape functions and the fine mesh given by local refinements of the coarse grid (refinement to \(h_i\)-\(p_i\) in the substructure \(\Omega_i)\). The preconditioner is constructed by two-level orthogonalization (elimination), and for the Poisson equation problem it is proved that the condition number for the preconditioned linear system is of the order \(\max_i (1+ \ln {H_ip_i \over h_i})^2\). This bound extends the well-known results for both the \(h\)-version \((p_i=1)\) and the \(p\)-version \((h_i= H_i)\).
      0 references
      \(h\)-\(p\) version
      0 references
      preconditioner
      0 references
      finite element method
      0 references
      Poisson equation
      0 references
      condition number
      0 references
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references