Higher order accurate discontinuous and continuous p-Galerkin methods for linear elastodynamics

Authors

  • T. Gleim,

    1. Institute of Mechanics and Dynamics, University of Kassel, Mönchebergstr. 7, 34109 Kassel, Germany
    Search for more papers by this author
    • Phone: +49 561 804 2698, Fax: +49 561 804 3631

  • D. Kuhl

    Corresponding author
    1. Institute of Mechanics and Dynamics, University of Kassel, Mönchebergstr. 7, 34109 Kassel, Germany
    • Institute of Mechanics and Dynamics, University of Kassel, Mönchebergstr. 7, 34109 Kassel, Germany
    Search for more papers by this author
    • Phone: +49 561 804 1815, Fax: +49 561 804 3631


Abstract

The present paper is concerned with the numerical integration of linear structural dynamics by means of different higher order accurate Galerkin time integration schemes. Firstly the single field discontinuous and continuous p-Galerkin schemes are developed in a generalized fashion using arbitrary polynomial degrees. In a further step the two field discontinuous and continuous p-Galerkin schemes are derived. In both algorithmic classes continuous Galerkin schemes are obtained by the strong enforcement of the continuity condition as special case of the discontinuous Galerkin schemes. The related time integration schemes are conditioned in an algorithmic set-up in such a manner that the implementation is similar to the classical Newmark scheme and its α derivates. Selected benchmark examples demonstrate the excellent dissipation and dispersion behavior and the robustness of the present Galerkin integration schemes. Furthermore, an error analysis, based on the analytical solution and the real local time integration error, verifies the order of accuracy of Galerkin time integration schemes controlled by the chosen polynomial degree. In particular the order of accuracy for each class of Galerkin integration schemes is specified. The comparison of the numerical effort for the classical Newmark scheme and the family of Galerkin schemes indicates advantages for Galerkin schemes: For a prescribed error level higher order Galerkin schemes are more effective than the Newmark scheme.

Ancillary