Stochastic homogenisation of free-discontinuity functionals in random perforated domains

Xavier Pellet,Lucia Scardia,Caterina Ida Zeppieri
DOI: https://doi.org/10.48550/arXiv.2002.01389
2020-02-05
Abstract:In this paper we study the asymptotic behaviour of a family of random free-discontinuity energies $E_\varepsilon$ defined on a randomly perforated domain, as $\varepsilon$ goes to zero. The functionals $E_\varepsilon$ model the energy associated to displacements of porous random materials that can develop cracks. To gain compactness for sequences of displacements with bounded energies, we need to overcome the lack of equi-coerciveness of the functionals. We do so by means of an extension result, under the assumption that the random perforations cannot come too close to one another. The limit energy is then obtained in two steps. As a first step we apply a general result of stochastic convergence of free-discontinuity functionals to a modified, coercive version of $E_\varepsilon$. Then the effective volume and surface energy densities are identified by means of a careful limit procedure.
Analysis of PDEs
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the stochastic homogenization problem of the free - discontinuous energy functional in randomly perforated domains. Specifically, the author studies the asymptotic behavior of the free - discontinuous energy functional \(E_{\varepsilon}\) defined on randomly perforated domains when the scale parameter \(\varepsilon\) tends to zero. These energy functionals \(E_{\varepsilon}\) describe the displacement energy of porous random materials when cracks occur. ### Main Problem Description 1. **Background and Motivation**: - **Free - discontinuous Problems**: These problems involve cracks or fractures in materials and are usually modeled by free - discontinuous energy functionals. - **Randomly Perforated Domains**: The research object is materials with randomly distributed holes, and the positions and sizes of these holes are all random. - **Stochastic Homogenization**: The goal is to understand the asymptotic behavior of the energy functional \(E_{\varepsilon}\) when the scale parameter \(\varepsilon\) tends to zero, that is, how to derive macroscopic properties from the microstructure. 2. **Specific Problems**: - **Definition of the Energy Functional**: \[ E_{\varepsilon}(\omega)(u, A)=\int_{A\setminus\varepsilon K(\omega)} f\left(\omega, \frac{x}{\varepsilon}, \nabla u\right)dx+\int_{S_{u}\cap(A\setminus\varepsilon K(\omega))} g\left(\omega, \frac{x}{\varepsilon}, \nu_{u}\right)dH^{n - 1} \] where \(u\in SBV(A)\), \(A\subset\mathbb{R}^{n}\) is a bounded Lipschitz domain, \(K(\omega)\) is a set of randomly distributed holes, and \(f\) and \(g\) are random variables representing volume and surface energy densities respectively. - **Main Challenges**: - **Lack of Equivalent Coercivity**: Due to the presence of holes, the energy functional \(E_{\varepsilon}\) lacks equivalent coercivity, which makes it difficult to ensure the compactness of the sequence. - **Natural Boundary Conditions**: The research assumes that natural boundary conditions are satisfied on the holes, which makes the compactness of the minimizing sequence more delicate. 3. **Solution Methods**: - **Extension Results**: By constructing an extension operator, the function \(u\) is extended from the perforated domain \(A\setminus\varepsilon K(\omega)\) to the entire domain \(A\) while maintaining the comparability of energy. - **Γ - Convergence**: Prove the Γ - convergence of the energy functional \(E_{\varepsilon}\) in the sense of \(L^{1}_{\text{loc}}(\mathbb{R}^{n})\) convergence to obtain the limit energy functional \(E_{\text{hom}}\). ### Main Results 1. **Compactness**: - For a fixed \(\omega\in\Omega\) and \(A\in A\), if the sequence \((u_{\varepsilon})\) satisfies \[ \sup_{\varepsilon > 0}\left(E_{\varepsilon}(\omega)(u_{\varepsilon}, A)+\|u_{\varepsilon}\|_{L^{\infty}(A\setminus\varepsilon K(\omega))}\right)<+\infty \] then there exists a sequence \((\tilde{u}_{\varepsilon})\)