Abstract:We consider divergence-based high order discretizations of an $L^2$-based first order system least squares formulation of a second order elliptic equation with Robin boundary conditions. For smooth geometries, we show optimal convergence rates in the $L^2(\Omega)$ norm for the scalar variable. Convergence rates for the $L^2(\Omega)$-norm error of the gradient of the scalar variable as well as vectorial variable are also derived. Numerical examples illustrate the analysis.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: in the second - order elliptic equation with non - homogeneous Robin boundary conditions, whether the first - kind system least - squares finite element method (FOSLS) based on the L2 norm can achieve the optimal convergence rate. Specifically, the author focuses on how to prove the optimal convergence rates of the scalar variable u, its gradient, and the vector variable φ in the L2(Ω) norm through high - order discretization methods under smooth geometric conditions.
### Specific problem description
1. **Model problem**:
The author considers a second - order elliptic equation with non - homogeneous Robin boundary conditions:
\[
-\Delta u+\gamma u = f\quad\text{in }\Omega,
\]
\[
\partial_n u+\alpha u = g\quad\text{on }\Gamma.
\]
where $\gamma,\alpha> 0$ are constants, $f\in L^2(\Omega)$ and $g\in L^2(\Gamma)$.
2. **First - kind system least - squares finite element method (FOSLS)**:
By introducing the vector variable $\phi = -\nabla u$, the above second - order equation is transformed into a first - order system:
\[
\nabla\cdot\phi+\gamma u = f\quad\text{in }\Omega,
\]
\[
\nabla u+\phi = 0\quad\text{in }\Omega,
\]
\[
\phi\cdot n-\alpha u=-g\quad\text{on }\Gamma.
\]
3. **Objectives**:
- Prove the optimal convergence rates of the scalar variable u, its gradient, and the vector variable φ in the L2 norm.
- Derive the error estimates of these variables in different norms, especially the estimates in the L2(Ω) and H(div,Ω) norms.
### Solutions
To achieve this goal, the author adopts the following methods:
- **Duality argument**: By introducing the dual problem and using the regularity of the dual solution to derive the required error estimates.
- **Approximation operator**: Develop a constrained approximation operator $I_h^\Gamma$ with specific orthogonal properties for error analysis.
- **Gradually improve estimates**: Systematically improve the error estimates of the scalar variable u, its gradient, and the vector variable φ by repeatedly using the duality argument.
### Main contributions
- **Optimal L2(Ω) convergence results**: For the scalar variable u, its gradient, and the trace in the least - squares approximation, explicit optimal L2(Ω) convergence results are obtained.
- **Improved vector variable estimates**: Derive improved L2(Ω) estimates of the vector variable φ, which explicitly depend on the mesh size h of the quasi - uniform mesh and the polynomial degree p.
- **Numerical verification**: Verify the theoretical convergence rates through numerical experiments and show the convergence behavior under finite Sobolev regularity.
In general, through strict mathematical analysis and numerical verification, this paper solves the problem of the optimal convergence rate of the FOSLS method in the second - order elliptic equation with non - homogeneous Robin boundary conditions.