Kunibert G. Siebert

Data Oscillation and Convergence of Adaptive FEM

Data oscillation is intrinsic information missed by the averaging process associated with finite element methods (FEM) regardless of quadrature. Ensuring a reduction rate of data oscillation, together with an error reduction based on a posteriori error estimators, we construct a simple and efficient adaptive FEM for elliptic partial differential equations (PDEs) with linear rate of convergence without any preliminary mesh adaptation nor explicit knowledge of constants. Any prescribed error tolerance is thus achieved in a finite number of steps. A number of numerical experiments in two and three dimensions yield quasi-optimal meshes along with a competitive performance.

Quasi-Optimal Convergence Rate for an Adaptive Finite Element Method

We analyze the simplest and most standard adaptive finite element method (AFEM), with any polynomial degree, for general second order linear, symmetric elliptic operators. As is customary in practice, the AFEM marks exclusively according to the error estimator and performs a minimal element refinement without the interior node property. We prove that the AFEM is a contraction, for the sum of the energy error and the scaled error estimator, between two consecutive adaptive loops. This geometric decay is instrumental to derive the optimal cardinality of the AFEM. We show that the AFEM yields a decay rate of the energy error plus oscillation in terms of the number of degrees of freedom as dictated by the best approximation for this combined nonlinear quantity.

Convergence of Adaptive Finite Element Methods

Adaptive finite element methods (FEMs) have been widely used in applications for over 20 years now. In practice, they converge starting from coarse grids, although no mathematical theory has been able to prove this assertion. Ensuring an error reduction rate based on a posteriori error estimators, together with a reduction rate of data oscillation (information missed by the underlying averaging process), we construct a simple and efficient adaptive FEM for elliptic partial differential equations. We prove that this algorithm converges with linear rate without any preliminary mesh adaptation nor explicit knowledge of constants. Any prescribed error tolerance is thus achieved in a finite number of steps. A number of numerical experiments in two and three dimensions yield quasi-optimal meshes along with a competitive performance. Extensions to higher order elements and applications to saddle point problems are discussed as well.

Theory of adaptive finite element methods: An introduction

This is a survey on the theory of adaptive finite element methods (AFEM), which are fundamental in modern computational science and engineering. We present a self-contained and up-to-date discussion of AFEM for linear second order elliptic partial differential equations (PDEs) and dimension d>1, with emphasis on the differences and advantages of AFEM over standard FEM. The material is organized in chapters with problems that extend and complement the theory. We start with the functional framework, inf-sup theory, and Petrov-Galerkin method, which are the basis of FEM. We next address four topics of essence in the theory of AFEM that cannot be found in one single article: mesh refinement by bisection, piecewise polynomial approximation in graded meshes, a posteriori error analysis, and convergence and optimal decay rates of AFEM. The first topic is of geometric and combinatorial nature, and describes bisection as a rather simple and efficient technique to create conforming graded meshes with optimal complexity. The second topic explores the potentials of FEM to compensate singular behavior with local resolution and so reach optimal error decay. This theory, although insightful, is insufficient to deal with PDEs since it relies on knowing the exact solution. The third topic provides the missing link, namely a posteriori error estimators, which hinge exclusively on accessible data: we restrict ourselves to the simplest residual-type estimators and present a complete discussion of upper and lower bounds, along with the concept of oscillation and its critical role. The fourth topic refers to the convergence of adaptive loops and its comparison with quasi-uniform refinement. We first show, under rather modest assumptions on the problem class and AFEM, convergence in the natural norm associated to the variational formulation. We next restrict the problem class to coercive symmetric bilinear forms, and show that AFEM is a contraction for a suitable error notion involving the induced energy norm. This property is then instrumental to prove optimal cardinality of AFEM for a class of singular functions, for which the standard FEM is suboptimal.

Local problems on stars: a posteriori error estimators, convergence, and performance

A new computable a posteriori error estimator is introduced, which relies on the solution of small discrete problems on stars. It exhibits built-in flux equilibration and is equivalent to the energy error up to data oscillation without any saturation assumption. A simple adaptive strategy is designed, which simultaneously reduces error and data oscillation, and is shown to converge without mesh pre-adaptation nor explicit knowledge of constants. Numerical experiments reveal a competitive performance, show extremely good effectivity indices, and yield quasi-optimal meshes.

Pointwise a posteriori error control for elliptic obstacle problems

AbstractWe consider a finite element method for the elliptic obstacle problem over polyhedral domains in ℝd, which enforces the unilateral constraint solely at the nodes. We derive novel optimal upper and lower a posteriori error bounds in the maximum norm irrespective of mesh fineness and the regularity of the obstacle, which is just assumed to be Hölder continuous. They exhibit optimal order and localization to the non-contact set. We illustrate these results with simulations in 2d and 3d showing the impact of localization in mesh grading within the contact set along with quasi-optimal meshes.

Fully Localized A posteriori Error Estimators and Barrier Sets for Contact Problems

We derive novel pointwise a posteriori error estimators for elliptic obstacle problems which, except for obstacle resolution, completely vanish within the full-contact set (localization). We then construct a posteriori barrier sets for free boundaries under a natural stability (or nondegeneracy) condition. We illustrate localization properties as well as reliability and efficiency for both solutions and free boundaries via several simulations in 2 and 3 dimensions.

A Unilaterally Constrained Quadratic Minimization with Adaptive Finite Elements

We consider obstacle problems where a quadratic functional associated with the Laplacian is minimized in the set of functions above a possibly discontinuous and thin but piecewise affine obstacle. In order to approximate minimum point and value, we propose an adaptive algorithm that relies on minima with respect to admissible linear finite element functions and on an a posteriori estimator for the error in the minimum value. It is proven that the generated sequence of approximate minima converges to the exact one. Furthermore, our numerical results in two and three dimensions indicate that the convergence rate with respect to the number of degrees of freedom is optimal in that it coincides with the one of nonlinear or adaptive approximation.

A Posteriori Error Analysis of Optimal Control Problems with Control Constraints

We derive a unifying framework for the a posteriori error analysis of control constrained linear-quadratic optimal control problems. We consider finite element discretizations with discretized and nondiscretized control. A fundamental error equivalence drastically simplifies the a posteriori error analysis for optimal control problems. It basically remains to apply error estimators for the linear state and adjoint problem. We give several examples, including stabilized discretizations, and investigate the quality of the estimators and the performance of the adaptive iteration by selected numerical experiments.

Pointwise a posteriori error estimates for monotone semi-linear equations

We derive upper and lower a posteriori estimates for the maximum norm error in finite element solutions of monotone semi-linear equations. The estimates hold for Lagrange elements of any fixed order, non-smooth nonlinearities, and take numerical integration into account. The proof hinges on constructing continuous barrier functions by correcting the discrete solution appropriately, and then applying the continuous maximum principle; no geometric mesh constraints are thus required. Numerical experiments illustrate reliability and efficiency properties of the corresponding estimators and investigate the performance of the resulting adaptive algorithms in terms of the polynomial order and quadrature.