Abstract:This work introduces finite element methods for a class of elliptic fully nonlinear partial differential equations. They are based on a minimal residual principle that builds upon the Alexandrov--Bakelman--Pucci estimate. Under rather general structural assumptions on the operator, convergence of $C^1$ conforming and discontinuous Galerkin methods is proven in the $L^\infty$ norm. Numerical experiments on the performance of adaptive mesh refinement driven by local information of the residual in two and three space dimensions are provided.
What problem does this paper attempt to address?
This paper aims to solve the numerical approximation problem of a class of fully nonlinear elliptic partial differential equations (PDEs). Specifically, the author introduced a finite - element method based on the least - residual principle to solve this type of equations, which are in the form:
\[ F(x, u, \nabla u, D^2 u) = f \quad \text{in} \quad \Omega \]
where \( \Omega \subset \mathbb{R}^n \) is a bounded polyhedral Lipschitz domain, \( f \in L^n(\Omega) \), and \( u = g \) on the boundary \( \partial \Omega \).
### Main problems:
1. **Numerical approximation**: How to design effective numerical methods to approximate the strong solution \( u \in C(\Omega) \cap W^{2,n}_{\text{loc}}(\Omega) \) of the above form.
2. **Convergence**: Ensure that the proposed finite - element method converges in the \( L^\infty \) norm.
3. **Adaptive mesh refinement**: Provide an adaptive mesh refinement algorithm based on local residual information to improve computational efficiency and accuracy.
### Key challenges:
- **PDEs in non - divergence form**: Different from common PDEs in divergence form, the treatment of this type of equations is more challenging.
- **Non - monotone methods**: In order to obtain a higher approximation order and a more flexible mesh refinement strategy, non - monotone methods need to be designed.
- **Non - smooth domains**: Allow consideration of non - smooth domains with corner singularities.
### Solutions:
The author proposed a least - residual method based on the Alexandrov–Bakelman–Pucci (ABP) estimate, which can ensure convergence under weaker regularity assumptions. Specific steps include:
- **Stability estimate**: Establish a stability estimate through the ABP maximum principle.
- **Conforming and non - conforming discretizations**: Study the convergence of \( C^1 \)-conforming and discontinuous Galerkin methods respectively.
- **A posteriori error estimate**: Use the stability estimate to construct a posteriori error estimates and design an adaptive mesh refinement algorithm accordingly.
### Application examples:
The paper also discussed several specific equations, such as linear PDEs, Pucci extremal operators, Hamilton - Jacobi - Bellman equations, and regularized Monge - Ampère equations, and proved that these equations satisfy the proposed theoretical framework.
In conclusion, this paper is committed to developing efficient and stable numerical methods for fully nonlinear elliptic PDEs, and provides theoretical basis and verification of practical applications.