Homogenized discontinuous Galerkin method for Maxwell's equations in periodic structured dispersive media (Q2869175)

From MaRDI portal





scientific article; zbMATH DE number 6242468
Language Label Description Also known as
default for all languages
No label defined
    English
    Homogenized discontinuous Galerkin method for Maxwell's equations in periodic structured dispersive media
    scientific article; zbMATH DE number 6242468

      Statements

      0 references
      0 references
      3 January 2014
      0 references
      Maxwell equations
      0 references
      dispersion
      0 references
      finite element method
      0 references
      discontinuous Galerkin method
      0 references
      homogenisation
      0 references
      multiscale asymptotic expansion
      0 references
      Homogenized discontinuous Galerkin method for Maxwell's equations in periodic structured dispersive media (English)
      0 references
      \textit{Y. Huang} et al. [J. Comput. Phys. 230, No. 12, 4559--4570 (2011; Zbl 1220.78015)] considered a discontinuous Galerkin (DG) method for solving the time-dependent Maxwell's equations in dispersive media.NEWLINENEWLINEIn the present paper, the authors extend this approach to dispersive media with periodic microstructures in space. A homogenisation technique yields a multiscale asymptotic expansion consisting of a space-dependent corrector function and a time-space-dependent electric field. Firstly, a DG method based on a bilinear form with a penalty term is applied to compute the corrector function in a unit cell. Secondly, a homogenised coefficient matrix can be determined by well-known methods. Thirdly, a partial differential equation inluding the homogenised coefficients yields the electric field. This equation is solved by a DG method in space again and a discretisation in time. The authors derive the formulas of their approach in three space dimensions, where the two-dimensional problem appears as a special case.NEWLINENEWLINEFinally, numerical simulations are presented for three test examples of two-dimensional problems: a time-independent equation, a computation of homogenised coefficients only and a full problem in time as well as in space. The results confirm that the approach is feasible and indicate the convergence of the method.NEWLINENEWLINEFor the entire collection see [Zbl 1264.65002].
      0 references
      0 references

      Identifiers

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