Received: 18 July 2016 Revised: 1 August 2017 Accepted: 30 April 2018
DOI: 10.1002/nme.5836
RESEARCH ARTICLE
On continuous, discontinuous, mixed, and primal hybrid
finite element methods for second-order elliptic problems
P. R. B. Devloo
1
C. O. Faria
2
A. M. Farias
3
S. M. Gomes
4
A. F. D. Loula
5
S. M. C. Malta
6
1
Faculdade de Engenharia Civil,
Universidade Estadual de Campinas,
Campinas, Brazil
2
Instituto de Matemática e Estatística,
Universidade do Estado do Rio de Janeiro,
Rio de Janeiro, Brazil
3
Departamento de Matemática, Instituto
Federal do Norte de Minas Gerais, Salinas,
Brazil
4
Instituto de Matemática, Estatística e
Computação Científica, Universidade
Estadual de Campinas, Campinas, Brazil
5
Coordenação de Modelagem
Computacional, Laboratório Nacional de
Computação Científica, Petrópolis, Brazil
6
Coordenação de Métodos Matemáticos e
Computacionais, Laboratório Nacional de
Computação Científica, Petrópolis, Brazil
Correspondence
Cristiane O. Faria, ANMAT, Instituto de
Matemática e Estatística, Universidade do
Estado do Rio de Janeiro, 20550-013
Rio de Janeiro, Brazil.
Email: cofaria@ime.uerj.br
Funding information
CNPq, the Brazilian Research Council,
Grant/Award Number: 310369/2006-1,
308632/2006-0, 352991/92-5, and
475259/2013-0
Summary
Finite element formulations for second-order elliptic problems, including the
classic H
1
-conforming Galerkin method, dual mixed methods, a discontinuous
Galerkin method, and two primal hybrid methods, are implemented and numer-
ically compared on accuracy and computational performance. Excepting the
discontinuous Galerkin formulation, all the other formulations allow static con-
densation at the element level, aiming at reducing the size of the global system
of equations. For a three-dimensional test problem with smooth solution, the
simulations are performed with h-refinement, for hexahedral and tetrahedral
meshes, and uniform polynomial degree distribution up to four. For a singu-
lar two-dimensional problem, the results are for approximation spaces based on
given sets of hp-refined quadrilateral and triangular meshes adapted to an inter-
nal layer. The different formulations are compared in terms of L
2
-convergence
rates of the approximation errors for the solution and its gradient, number of
degrees of freedom, both with and without static condensation. Some insights
into the required computational effort for each simulation are also given.
KEYWORDS
adaptivity, continuous Galerkin, discontinuous Galerkin, finite element methods, hybrid
formulations, mixed formulation
1 INTRODUCTION
This work intends to be a survey on six finite element methods for second-order elliptic problems. A comparison study is
presented in terms of stability, accuracy, flexibility in the choice of approximation spaces, solution strategy, and also some
aspects of computational efficiency. The considered formulations use approximation spaces according to their original
weak forms as follows:
1. single field formulations (discontinuous Galerkin (DG) and continuous Galerkin methods),
Int J Numer Methods Eng. 2018;1–25. wileyonlinelibrary.com/journal/nme Copyright © 2018 John Wiley & Sons, Ltd. 1