Monolithic two-level Schwarz preconditioner for Biot's consolidation model in two space dimensions

Stefan Meggendorfer,Guido Kanschat,Johannes Kraus
2024-04-25
Abstract:This paper addresses the construction and analysis of a class of domain decomposition methods for the iterative solution of the quasi-static Biot problem in three-field formulation. The considered discrete model arises from time discretization by the implicit Euler method and space discretization by a family of strongly mass-conserving methods exploiting $H^{div}$-conforming approximations of the solid displacement and fluid flux fields. For the resulting saddle-point problem, we construct monolithic overlapping domain decomposition (DD) methods whose analysis relies on a transformation into an equivalent symmetric positive definite system and on stable decompositions of the involved finite element spaces under proper problem-dependent norms. Numerical results on two-dimensional test problems are in accordance with the provided theoretical uniform convergence estimates for the two-level multiplicative Schwarz method.
Numerical Analysis
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to construct and analyze a class of domain decomposition methods for iterative solutions to quasi - static Biot problems. Specifically, the author designed an overall two - level Schwarz preconditioning method for the three - field (solid - phase displacement, fluid flux, pore pressure) discrete model. This method relies on transforming the saddle - point problem into an equivalent singularly perturbed symmetric positive - definite problem and performing a stable decomposition of the finite - element space under an appropriate norm. ### Analysis of the Core Problems in the Paper 1. **Problem Background**: - Biot's consolidation model describes the interaction between fluid flow and mechanical deformation in porous media. - This model is widely used in fields such as petroleum engineering, carbon dioxide sequestration, and biomechanics of biological tissues. 2. **Research Objectives**: - To construct an effective domain decomposition method for solving the saddle - point problem in the Biot model. - The proposed method should be parameter - robust, that is, its performance is not affected by changes in model parameters (such as permeability, Lamé parameters, etc.). 3. **Solutions**: - Use the implicit Euler method for time discretization and the mixed finite - element method with strong mass conservation for space discretization. - Design an overall two - level Schwarz method based on overlapping vertex patches, which is suitable for the three - field Biot model. - Prove the convergence of the method by transforming the saddle - point problem into an equivalent singularly perturbed symmetric positive - definite problem. 4. **Theoretical Contributions**: - Provide theoretically consistent convergence estimates for the two - level multiplicative Schwarz method. - Numerical experiments verify the effectiveness of the theoretical results. ### Formula Summary - **Biot Equations**: \[ \begin{aligned} &-\text{div}(2\mu \varepsilon(u) + \lambda \text{div}(u)I) + \alpha \nabla p = f \quad &\text{in } \Omega\times(0, T), \\ &v + K \nabla p = 0 \quad &\text{in } \Omega\times(0, T), \\ &-\alpha \text{div}(\partial_t u) - \text{div}(v) - c_s \partial_t p = g \quad &\text{in } \Omega\times(0, T), \end{aligned} \] where: - \( \mu \) and \( \lambda \) are Lamé coefficients of linearly elastic materials, - \( \alpha \) is the Biot - Willis constant, - \( K \) is a symmetric positive - definite permeability tensor, - \( c_s \) is the specific storage coefficient. - **Discrete System**: \[ \begin{aligned} &-2\mu \text{div}(\varepsilon(u)) - \lambda \nabla (\text{div}(u)) + \alpha \nabla p = f \quad &\text{in } \Omega, \\ &v + K \nabla p = 0 \quad &\text{in } \Omega, \\ &-\frac{\alpha}{\tau} \text{div}(u) - \text{div}(v) - \frac{c_s}{\tau} p = \bar{g} \quad &\text{in } \Omega, \end{aligned} \] where \( \bar{g}=g - \frac{\alpha}{\tau} \text{div}(u_{\text{old}})-\frac{c_s}{\tau} p_{\text{old}} \), \( u_{\text{old}} \)