Abstract:This paper deals with stability of classical Runge-Kutta collocation methods. When such methods are embedded in linearly implicit methods as developed in [12] and used in [13] for the time integration of nonlinear evolution PDEs, the stability of these methods has to be adapted to this context. For this reason, we develop in this paper several notions of stability, that we analyze. We provide sufficient conditions that can be checked algorithmically to ensure that these stability notions are fulfilled by a given Runge-Kutta collocation method. We also introduce examples and counterexamples used in [13] to highlight the necessity of these stability conditions in this context.
What problem does this paper attempt to address?
### Problems the paper attempts to solve
The paper aims to explore the stability issues of the classical Runge - Kutta collocation methods in a specific context. Specifically, when these methods are embedded into linear implicit methods for the time integration of nonlinear evolution partial differential equations (PDEs), their stability needs to adapt to this new context. To achieve this goal, the paper introduces and analyzes multiple stability concepts and provides sufficient conditions that can be algorithmically verified to ensure that a given Runge - Kutta collocation method satisfies these stability conditions. In addition, the paper also demonstrates the necessity of these stability conditions in this context through examples and counter - examples.
### Main research contents
1. **Introduction and analysis of stability concepts**:
- The author introduces several stability concepts, especially suitable for the stability analysis of Runge - Kutta collocation methods in linear implicit methods.
- For dissipative problems (such as the heat equation), \(\hat{A}\)-stability is introduced.
- For oscillatory problems (such as the Schrödinger equation), \(\hat{I}\)-stability is introduced.
2. **Sufficient conditions for stability conditions**:
- Sufficient conditions that the coefficients of the Runge - Kutta collocation method should satisfy are provided to ensure that these stability concepts hold.
- These conditions can be algorithmically verified, for example, by using the Routh - Hurwitz algorithm.
3. **Examples and counter - examples**:
- Through specific examples and counter - examples, the importance and necessity of these stability conditions in practical applications are demonstrated.
### Specific definitions of stability concepts
- **A - stability**: For all \(\lambda\in\mathbb{C}^-\), \(|R(\lambda)|\leq1\).
- **I - stability**: For all \(\lambda\in i\mathbb{R}\), \(|R(\lambda)|\leq1\).
- **AS - stability**: For all \(\lambda\in\mathbb{C}^-\), the matrix \((I - \lambda A)\) is invertible and \(\lambda\mapsto\lambda b^T(I - \lambda A)^{-1}\) is uniformly bounded on \(\mathbb{C}^-\).
- **ASI - stability**: For all \(\lambda\in\mathbb{C}^-\), the matrix \((I - \lambda A)\) is invertible and \(\lambda\mapsto(I - \lambda A)^{-1}\) is uniformly bounded on \(\mathbb{C}^-\).
- **IS - stability**: For all \(\lambda\in i\mathbb{R}\), the matrix \((I - \lambda A)\) is invertible and \(\lambda\mapsto\lambda b^T(I - \lambda A)^{-1}\) is uniformly bounded on \(i\mathbb{R}\).
- **ISI - stability**: For all \(\lambda\in i\mathbb{R}\), the matrix \((I - \lambda A)\) is invertible and \(\lambda\mapsto(I - \lambda A)^{-1}\) is uniformly bounded on \(i\mathbb{R}\).
- **\(\hat{A}\)-stability**: Satisfy A - stability, AS - stability and ASI - stability simultaneously.
- **\(\hat{I}\)-stability**: Satisfy I - stability, IS - stability and ISI - stability simultaneously.
### Conclusion
The paper provides sufficient conditions to ensure the stability of the Runge - Kutta collocation method in a specific context by introducing and analyzing multiple stability concepts. These conditions can be algorithmically verified, thus providing theoretical support for practical applications. In addition, through examples and counter - examples.