Alan F. Hegarty,Eugene O'Riordan
Abstract:Fitted finite element methods are constructed for a singularly perturbed convection-diffusion problem in two space dimensions. Exponential splines as basis functions are combined with Shishkin meshes to obtain a stable parameter-uniform numerical method. These schemes satisfy a discrete maximum principle. In the classical case, the numerical approximations converge, in the maximum pointwise norm, at a rate of second order and the approximations converge at a rate of first order for all values of the singular perturbation parameter.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the numerical computation difficulties encountered when solving the singularly perturbed convection - diffusion problem in two - dimensional space. Specifically, such problems are prone to spurious oscillations during numerical solution, and these oscillations can seriously affect the stability and accuracy of the solution. To overcome this challenge, the author proposes to use exponential splines as basis functions and combine Shishkin meshes to construct a stable parameter - uniform numerical method. This method can satisfy the discrete maximum principle and can achieve a first - order convergence rate for all singularly perturbed parameter values under the classical \(L_{\infty}\) norm; and in the non - singularly perturbed case, a second - order convergence rate can be achieved.
### Core problems of the paper
- **Spurious oscillations**: The convection - diffusion problem is prone to generate spurious oscillations during numerical solution, especially in the boundary layer region.
- **Parameter uniformity**: It is necessary to construct a numerical method so that its error bound does not depend on the singularly perturbed parameter \(\varepsilon\).
- **Stability**: The numerical method needs to satisfy the discrete maximum principle to ensure the stability of the solution.
### Solutions
- **Exponential splines**: Use exponential splines as basis functions, and these spline functions can better capture the boundary layer behavior.
- **Shishkin meshes**: Adopt Shishkin meshes, which are densified in the boundary layer region, thereby improving the accuracy of the solution.
- **Discrete maximum principle**: The constructed numerical method satisfies the discrete maximum principle, ensuring the stability of the solution and the absence of spurious oscillations.
### Mathematical model
Consider the singularly perturbed convection - diffusion problem in the following form:
\[
-\varepsilon\Delta u+\vec{a}\cdot\nabla u + bu = f,\quad(x,y)\in\Omega:=(0,1)^{2}
\]
\[
u = 0,\quad(x,y)\in\partial\Omega
\]
where \(\vec{a}=(a_{1},a_{2})\), and \(a_{1}>\alpha_{1}>0\), \(a_{2}>\alpha_{2}>0\), \(b\geq0\), \(f\in C^{1,\gamma}(\Omega)\).
### Numerical method
- **Finite element framework**: Use exponential splines as basis functions and combine Shishkin meshes for discretization.
- **Discrete weak form**: By introducing an appropriate discrete weak form, construct a numerical solution \(U\) such that
\[
B(U,V)=(f,V),\quad\forall V\in T_{N}
\]
where \(B(U,V)=\varepsilon(\nabla U,\nabla V)+(\vec{a}\cdot\nabla U,V)+(bU,V)\).
### Convergence analysis
- **Node convergence**: At the nodes, the error bound of the numerical solution \(U\) is
\[
|(U - u)(x_{i},y_{j})|\leq CN^{-1},\quad(x_{i},y_{j})\in\Omega_{N}
\]
- **Global convergence**: Globally, the error bound of the numerical solution \(U\) is
\[
\|U - u\|_{\infty}\leq CN^{-1}
\]
In the non - singularly perturbed case (\(\varepsilon\geq\varepsilon_{0}>0\)), the error bound can reach second - order convergence:
\[
\|U - u\|_{\infty}\leq