Abstract
We establish the existence of liftings into discrete subspaces of \(\varvec{H}({{\mathrm{div}}})\) of piecewise polynomial data on locally refined simplicial partitions of polygonal/polyhedral domains. Our liftings are robust with respect to the polynomial degree. This result has important applications in the a posteriori error analysis of parabolic problems, where it permits the removal of so-called transition conditions that link two consecutive meshes. It can also be used in the a posteriori error analysis of elliptic problems, where it allows the treatment of meshes with arbitrary numbers of hanging nodes between elements. We present a constructive proof based on the a posteriori error analysis of an auxiliary elliptic problem with \(H^{-1}\) source terms, thereby yielding results of independent interest. In particular, for such problems, we obtain guaranteed upper bounds on the error along with polynomial-degree robust local efficiency of the estimators.
Similar content being viewed by others
References
Agnelli, J.P., Garau, E.M., Morin, P.: A posteriori error estimates for elliptic problems with Dirac measure terms in weighted spaces. ESAIM Math. Model. Numer. Anal. 48, 1557–1581 (2014)
Ainsworth, M.: A framework for obtaining guaranteed error bounds for finite element approximations. J. Comput. Appl. Math. 234, 2618–2632 (2010)
Ainsworth, M., Guzmán, J., Sayas, F.-J.: Discrete extension operators for mixed finite element spaces on locally refined meshes. Math. Comp. 85, 2639–2650 (2016)
Araya, R., Behrens, E., Rodríguez, R.: A posteriori error estimates for elliptic problems with Dirac delta source terms. Numer. Math. 105, 193–216 (2006)
Braess, D., Pillwein, V., Schöberl, J.: Equilibrated residual error estimates are \(p\)-robust. Comput. Methods Appl. Mech. Engrg. 198, 1189–1197 (2009)
Braess, D., Schöberl, J.: Equilibrated residual error estimator for edge elements. Math. Comp. 77, 651–672 (2008)
Cohen, A., DeVore, R., Nochetto, R.H.: Convergence rates of AFEM with \(H^{-1}\) data. Found. Comput. Math. 12, 671–718 (2012)
Costabel, M., McIntosh, A.: On Bogovskiĭ and regularized Poincaré integral operators for de Rham complexes on Lipschitz domains. Math. Z. 265, 297–320 (2010)
Demkowicz, L., Gopalakrishnan, J., Schöberl, J.: Polynomial extension operators. Part III. Math. Comp. 81, 1289–1326 (2012)
Destuynder, P., Métivet, B.: Explicit error bounds in a conforming finite element method. Math. Comp. 68, 1379–1396 (1999)
Dolejší, V., Ern, A., Vohralík, M.: hp-adaptation driven by polynomial-degree-robust a posteriori error estimates for elliptic problems. SIAM J. Sci. Comput. 38(5), A3220–A3246 (2016)
Ern, A., Vohralík, M.: Polynomial-degree-robust a posteriori estimates in a unified setting for conforming, nonconforming, discontinuous Galerkin, and mixed discretizations. SIAM J. Numer. Anal. 53, 1058–1081 (2015)
Ern, A., Vohralík, M.: Stable broken \(H^1\) and H(div) polynomial extensions for polynomial-degree-robust potential and flux reconstruction in three space dimensions, submitted for publication (2016), https://hal.archives-ouvertes.fr/hal-01422204 (preprint)
Ern, A., Smears, I., Vohralík, M.: Guaranteed, locally space-time efficient, and polynomial-degree robust a posteriori error estimates for high-order discretizations of parabolic problems, submitted for publication (2016), https://hal.archives-ouvertes.fr/hal-01377086 (preprint)
Gaspoz, F., Morin, P., Veeser, A.: A posteriori error estimates with point sources in fractional Sobolev spaces. Numer. Methods Par. Differ. Equ. (2016). doi:10.1002/num.22065
Kreuzer, C., Veeser, A.: Private communication
Verfürth, R.: A posteriori error estimates for finite element discretizations of the heat equation. Calcolo 40, 195–212 (2003)
Author information
Authors and Affiliations
Corresponding author
Additional information
This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 647134 GATIPOR).
Application to a posteriori error estimates on meshes with hanging nodes
Application to a posteriori error estimates on meshes with hanging nodes
Equilibrated flux a posteriori error estimates for meshes with hanging nodes are developed in [11] where the equilibration is performed on patches \(\omega _{\mathbf {a}}\) corresponding to the support of hat functions \(\psi _{\mathbf {a}}\) associated with non-hanging nodes \(\mathbf {a}\) of the computational mesh and forming a partition of unity of the computational domain, see [11, Assumption 2.1]. It turns out that a slight extension of the equilibration patch \(\omega _{\mathbf {a}}\) of [11] enables the removal of the usual dependence of a posteriori efficiency constants on the number of levels of hanging nodes, thereby allowing for a completely arbitrary number of levels of hanging nodes. More precisely, it suffices to extend the equilibration patch \(\omega _{\mathbf {a}}\) so that all the products \(h_{\omega _{\mathbf {a}}} ||\nabla \psi _{\mathbf {a}}||_{\infty , \omega _{\mathbf {a}}}\) are uniformly bounded. Then, applying Theorem 1.2 where the patch \(\omega _{\mathbf {a}}\) of [11] corresponds here to the domain \(\varOmega \), and the hat function \(\psi _{\mathbf {a}}\) of [11] corresponds here to the function \(\psi _{\dagger }\), we infer that the factor \(h_{\omega _{\mathbf {a}}} \max _{\hat{\mathbf {a}} \in \widehat{\mathcal {V}}_{\mathbf {a}}} ||\nabla \psi _{\hat{\mathbf {a}}}||_{\infty ,\omega _{\hat{\mathbf {a}}}}\) of [11, Theorem 3.12] can be replaced by the factor \(h_{\omega _{\mathbf {a}}} ||\nabla \psi _{\mathbf {a}}||_{\infty , \omega _{\mathbf {a}}}\), i.e. \(h_{\varOmega }||\nabla \psi _{\dagger }||_{\infty }\) in the present notation. The extension of the equilibration patch is illustrated in Fig. 1. This extension typically entails including several layers of fine elements, so as to ensure that the factors \(h_{\omega _{\mathbf {a}}} ||\nabla \psi _{\mathbf {a}}||_{\infty , \omega _{\mathbf {a}}}\) are uniformly bounded. The price to pay to achieve robustness with respect to the level of hanging nodes is thus a somewhat more expensive computation of the equilibrated flux. The proof of Theorem 1.2 in Sect. 6 shows that this cost can be significantly reduced to the solution of two low-order systems over the extended patch, followed by local high-order corrections within the extended patch.
Rights and permissions
About this article
Cite this article
Ern, A., Smears, I. & Vohralík, M. Discrete p-robust \(\varvec{H}({{\mathrm{div}}})\)-liftings and a posteriori estimates for elliptic problems with \(H^{-1}\) source terms. Calcolo 54, 1009–1025 (2017). https://doi.org/10.1007/s10092-017-0217-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10092-017-0217-4