Abstract:Finite Difference methods (FD) are one of the oldest and simplest methods for solving partial differential equations (PDE). Block Finite Difference methods (BFD) are FD methods in which the domain is divided into blocks, or cells, containing two or more grid points, with a different scheme used for each grid point, unlike the standard FD method.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to improve the accuracy and stability of the finite difference method (FD) in solving partial differential equations (PDEs), especially by introducing the error inhibition phenomenon to achieve a global error order higher than the truncation error order. Specifically:
1. **Error Inhibition Phenomenon**: In the traditional finite difference method, the global error order is the same as the truncation error order. However, the paper explores how to make the global error order higher than the truncation error order through the block finite difference method (BFD), thereby obtaining a higher convergence rate.
2. **Relationship between BFD Method and DG Method**: The paper shows that the BFD method can be regarded as a form of the discontinuous Galerkin (DG) method, and uses the stability analysis tools of the DG method to prove the stability of the BFD method.
3. **Treatment of Non - periodic Boundary Conditions**: The paper also discusses how to apply the BFD method to non - periodic problems with Dirichlet boundary conditions, and verifies its stability and convergence through numerical experiments.
4. **Two - Dimensional Generalization**: The paper further generalizes the above - mentioned methods to the two - dimensional case, including periodic and non - periodic boundary conditions, and verifies the correctness of the theoretical analysis through numerical results.
In summary, this paper aims to improve the accuracy and stability of numerical solutions by improving the finite difference method, especially in terms of error inhibition, and to provide effective numerical solution schemes for complex geometries and higher - dimensional problems.
### Formula Examples
- **Definition of Truncation Error**:
\[
(Te)_j = P(w(x_j))-(Qw)_j
\]
where \(w(x)\) is a smooth function, and \(w\) is the projection of \(w(x)\) on the grid.
- **Definition of Global Error**:
\[
E = u - v
\]
where \(u\) is the projection of \(u(x,t)\) on the grid, and \(v\) is the numerical solution.
- **Error Evolution Equation**:
\[
\frac{\partial E}{\partial t}=QE + Te
\]
- **Stability Condition**:
\[
\Theta_{j - 1/2}=\theta_{j - 1/2}+\frac{1}{2}\int_{x_{j - 3/2}}^{x_{j - 1/2}}(\partial_x u)^2dx+\frac{1}{2}\int_{x_{j - 1/2}}^{x_{j + 1/2}}(\partial_x u)^2dx
\]
where \(\theta_{j - 1/2}\) is the energy contribution at the boundary.
These formulas show the key mathematical expressions involved in the paper, ensuring the correctness and readability of the formulas.