Abstract
We prove a compactness and semicontinuity result that applies to minimisation problems in nonhomogeneous linear elasticity under Dirichlet boundary conditions. This generalises a previous compactness theorem that we proved and employed to show existence of minimisers for the Dirichlet problem for the (homogeneous) Griffith energy.
1. Introduction
Predicting how non-uniform elastic materials—such as composites or layered structures—crack and deform under stress is a fundamental challenge in fracture mechanics. This paper, authored by Antonin Chambolle and Vito Crismale, provides the mathematical rigorous foundation needed to predict these behaviors. By proving "compactness" and "lower semicontinuity" results, the authors demonstrate that energy-minimizing equilibrium states actually exist for these complex systems.
In the framework of brittle fracture, the equilibrium configuration of a solid is typically found by minimizing a total energy functional. This functional typically consists of two parts: the bulk elastic energy (how much the material stretches) and the surface energy (the energy cost to create a crack). While existence results for homogeneous (uniform) materials were established using the Griffith energy model, real-world materials are rarely uniform. When material properties vary with position ($x$), the mathematics becomes significantly more difficult. A major "pathological" issue arises: if a piece of the material breaks off completely, it can undergo a "rigid motion" (sliding or rotating) without changing its internal elastic energy. In a non-uniform material, this can lead to sequences of configurations where the displacement blows up to infinity, making it hard to prove that a stable "limit" state exists.
2. Mathematical Framework and Preliminaries
The authors work within the framework of Calculus of Variations, specifically using the space of Generalized Special functions of Bounded Deformation ($GSBD^p$). This space is designed to handle functions that have "jumps," which represent physical cracks.
The core of their approach involves several advanced mathematical techniques:
- Slicing Arguments: To analyze the surface energy (the cracks), they use a technique that reduces the $n$-dimensional problem into 1D "slices," allowing them to apply simpler one-dimensional results to the complex multi-dimensional jump set.
- Korn-Poincaré Inequalities: They utilize advanced versions of these inequalities tailored for functions with small jump sets to control the displacement field based on the elastic strain.
3. The Compactness Result
The most significant insight of the paper is that the lack of compactness in nonhomogeneous problems isn't a dead end. The authors introduce a method to restore compactness by accounting for the "drift" of fractured components.
Theorem 1.1 (Compactness): This theorem proves that any sequence of displacements with bounded energy has a subsequence that, after adjusting for piecewise rigid motions, converges to a limit in the $GSBD^p$ space. By partitioning the material into a "Caccioppoli partition" (a mathematical way of defining fractured chunks) and assigning each chunk its own rigid motion (a different rotation/translation for each fractured piece), the authors show the system still behaves predictably in the limit.
4. Lower Semicontinuity and Minimisation
The "holy grail" for this type of problem is establishing the existence of minimizers, which requires proving the lower semicontinuity of the energy functional. This ensures that the limit state doesn't suddenly have less energy than the states approaching it.
Theorem 1.2 (Existence of Minimizers): This result establishes the lower semicontinuity of the energy functional under Dirichlet boundary conditions (where the displacement is fixed at the boundaries). This directly implies that for a wide class of nonhomogeneous materials, an equilibrium configuration with cracks actually exists and can be found through minimization.
A crucial mathematical tool introduced here is the term $g_\infty$. This represents the surface energy density when the "jump" (the gap in the crack) becomes infinitely large. This ensures the functional is well-behaved even as cracks widen significantly. This work effectively generalizes the classic Griffith fracture energy—where the crack cost is simply proportional to the crack area—to cases where the cost of a crack can depend on its location, the direction of the crack, and the size of the opening.
5. Limitations and Future Work
The authors note that their condition for the surface energy at infinity ($g_\infty$) is currently restrictive. It assumes the energy becomes independent of the jump amplitude for very large cracks; they admit they currently have "no idea" how to treat more general cases where the energy might continue to evolve with the jump size. Furthermore, while these results provide a necessary starting point for a theory of crack evolution, the paper focuses on static equilibrium (static configurations) rather than the quasi-static evolution of cracks over time.
Summary updated