A Three-Field Multiscale Method

Franklin de Barros,Alexandre L. Madureira,Frédéric Valentin
2024-04-26
Abstract:"A Three-Field Domain Decomposition Method" is the title of a seminal paper by F. Brezzi and L. D. Marini which introduces a three-field formulation for elliptic partial differential equations. Based on that, we propose the Multiscale-Hybrid-Hybrid Method (MH$^2$M) for the Darcy model, a multiscale finite element method that yields, after a series of formal manipulations, a symmetric positive definite formulation that depends only on the trace of the solution. We show stability and convergence results for a family of finite element spaces and establish relationships with other multiscale finite element methods.
Numerical Analysis
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to simulate fluid - flow problems efficiently and accurately in a multi - scale environment, especially the flow problems in complex porous media encountered in fields such as oil reservoirs, pollutant transport, and water resource management. Specifically, the paper proposes a new method named Multiscale - Hybrid - Hybrid Method (MH2M), aiming to improve the existing multi - scale finite element methods to better handle the multi - scale coefficient problems in the Darcy model. ### Main contributions: 1. **Propose a new multi - scale method**: The MH2M method relaxes the continuity requirement of flow variables by introducing a second Lagrange multiplier and enforces its continuity through the weak form. This method results in a positive - definite symmetric global linear system instead of the traditional saddle - point problem. 2. **Improve the approximation of flux variables**: The MH2M method uses the Dirichlet - to - Neumann operator derived from the new local Neumann problem to construct multi - scale flux basis functions, which is more in line with physical characteristics than traditional methods. 3. **Achieve weak continuity**: On the coarse - grid skeleton, the MH2M method achieves weak continuity of discrete primal variables (pressure) and dual variables (flow velocity) while maintaining local equilibrium. 4. **Theoretical analysis**: Prove the well - posedness and optimal approximation properties of the MH2M method and propose a family of interpolation spaces that satisfy these conditions. 5. **Relations with existing methods**: Establish the relationship between the MH2M method and other multi - scale finite element methods (such as the MHM method and MsFEM). ### Mathematical background: The main equation involved in the paper is the Darcy equation, which is an elliptic partial differential equation used to describe fluid flow in porous media. The specific form is as follows: \[ -\text{div}(A \nabla u)=f \quad \text{in} \ \Omega \] \[ u = 0 \quad \text{in} \ \partial \Omega \] where \(\Omega \subset \mathbb{R}^d\) is an open bounded region, \(d = 2\) or \(3\), and the boundary \(\partial \Omega\) is a polyhedral Lipschitz boundary. The tensor \(A \in [L^\infty(\Omega)]^{d \times d}_{\text{sym}}\) is uniformly positive - definite, and for almost all \(x \in \Omega\), there exist positive constants \(a_{\text{min}}\) and \(a_{\text{max}}\) such that: \[ a_{\text{min}} |v|^2 \leq A(x) v \cdot v \leq a_{\text{max}} |v|^2 \quad \forall v \in \mathbb{R}^d \] ### Method overview: 1. **Decompose the exact solution**: Decompose the exact solution into local and global contribution parts. After discretization, the local and global problems can be treated separately. 2. **Construct multi - scale basis functions**: Construct multi - scale flux basis functions through the new local Neumann problem. 3. **Implement the weak form**: Introduce a second Lagrange multiplier to achieve weak continuity and construct a positive - definite symmetric global linear system. 4. **Numerical analysis**: Prove the well - posedness and optimal approximation properties of the method and give specific error estimates. In conclusion, this paper provides a new and efficient solution to fluid - flow problems in a multi - scale environment by proposing the MH2M method.