English

A Note on the Reliability of Goal-Oriented Error Estimates for Galerkin Finite Element Methods with Nonlinear Functionals

Numerical Analysis 2025-11-07 v2 Computational Engineering, Finance, and Science Numerical Analysis

Abstract

We consider estimating the discretization error in a nonlinear functional J(u)J(u) in the setting of an abstract variational problem: find uVu \in \mathcal{V} such that B(u,φ)=L(φ)  φVB(u,\varphi) = L(\varphi) \; \forall \varphi \in \mathcal{V}, as approximated by a Galerkin finite element method. Here, V\mathcal{V} is a Hilbert space, B(,)B(\cdot,\cdot) is a bilinear form, and L()L(\cdot) is a linear functional. We consider well-known error estimates η\eta of the form J(u)J(uh)η=L(z)B(uh,z)J(u) - J(u_h) \approx \eta = L(z) - B(u_h, z), where uhu_h denotes a finite element approximation to uu, and zz denotes the solution to an auxiliary adjoint variational problem. We show that there exist nonlinear functionals for which error estimates of this form are not reliable, even in the presence of an exact adjoint solution solution zz. An estimate η\eta is said to be reliable if there exists a constant CR>0C \in \mathbb{R}_{>0} independent of uhu_h such that J(u)J(uh)Cη|J(u) - J(u_h)| \leq C|\eta|. We present several example pairs of bilinear forms and nonlinear functionals where reliability of η\eta is not achieved.

Cite

@article{arxiv.2506.09913,
  title  = {A Note on the Reliability of Goal-Oriented Error Estimates for Galerkin Finite Element Methods with Nonlinear Functionals},
  author = {Brian N. Granzow and Stephen D. Bond and D. Thomas Seidl and Bernhard Endtmayer},
  journal= {arXiv preprint arXiv:2506.09913},
  year   = {2025}
}

Comments

6 pages, 1 figure, 1 table

R2 v1 2026-07-01T03:11:38.105Z