Abstract:In this paper, we use an implicit two-derivative deferred correction time discretization approach and combine it with a spatial discretization of the discontinuous Galerkin spectral element method to solve (non-)linear PDEs. The resulting numerical method is high order accurate in space and time. As the novel scheme handles two time derivatives, the spatial operator for both derivatives has to be defined. This results in an extended system matrix of the scheme. We analyze this matrix regarding possible simplifications and an efficient way to solve the arising (non-)linear system of equations. It is shown how a carefully designed preconditioner and a matrix-free approach allow for an efficient implementation and application of the novel scheme. For both, linear advection and the compressible Euler equations, up to eighth order of accuracy in time is shown. Finally, it is illustrated how the method can be used to approximate solutions to the compressible Navier-Stokes equations.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to develop an efficient and high - precision time - discretization method for solving nonlinear partial differential equations (PDEs). Specifically, the author combines the implicit second - order derivative deferred - correction time - discretization method with the spatial - discretization method of the Discontinuous Galerkin Spectral Element Method (DGSEM) and proposes a new numerical method, which has high - order accuracy in both space and time.
### Main Problems and Solutions
1. **Requirement for High - Order Accuracy**:
- **Problem**: Traditional single - derivative time - discretization methods (such as Runge - Kutta or BDF methods) can only use the first - order time derivative \( \frac{\partial w}{\partial t} \) for time - stepping, which limits the accuracy and efficiency of the method.
- **Solution**: Introduce the second - order derivative \( \frac{\partial^2 w}{\partial t^2} \), and through the implicit second - order derivative deferred - correction method, higher time accuracy can be achieved with fewer time nodes.
2. **Handling of Extended System Matrices**:
- **Problem**: Since two time derivatives need to be processed, the spatial operator also needs to define two derivatives, resulting in the expansion of the system matrix and an increase in computational complexity.
- **Solution**: Through carefully designed pre - conditioners and matrix - free methods, the author shows how to efficiently solve the extended nonlinear equations. In particular, using the Schur complement method can reduce the problem size, thereby reducing the complexity of calculation and implementation.
3. **Handling of Nonlinear PDEs**:
- **Problem**: Solving nonlinear PDEs (such as the compressible Euler equations and the Navier - Stokes equations) is usually more complex and requires efficient numerical methods to ensure stability and accuracy.
- **Solution**: The author shows how to apply the proposed new method to nonlinear PDEs and verifies its high - order accuracy in time.
### Mathematical Model
- **Basic Equation**:
\[
\frac{\partial w}{\partial t}+\nabla_x\cdot F(w) = 0
\]
where \( w \) is the state vector and \( F(w) \) is the flux function.
- **Time Derivative**:
\[
\frac{\partial^2 w}{\partial t^2}=\nabla_x\cdot\left(-\frac{\partial F(w)}{\partial w}\frac{\partial w}{\partial t}\right)
\]
- **Discretization**:
- Spatial discretization uses the Discontinuous Galerkin Spectral Element Method (DGSEM).
- Time discretization uses the implicit second - order derivative deferred - correction method.
### Experimental Results
- **Linear Advection Equation**:
- The convergence of the method under different time steps is verified, and the effects of using the BJ pre - conditioner and the Schur complement method are demonstrated.
- **Compressible Euler Equation**:
- The effectiveness and high - order accuracy of the method in the nonlinear case are verified through numerical experiments.
### Conclusion
This paper proposes a new numerical method that combines the implicit second - order derivative deferred - correction method and the Discontinuous Galerkin Spectral Element Method, and successfully solves the problems of high - order accuracy and solving nonlinear PDEs. Through carefully designed pre - conditioners and matrix - free methods, the computational efficiency and stability are improved.