Stability analysis of two-step Runge-Kutta methods for delay differential equations (Q1609162): Difference between revisions

This paper is concerned with the so called P-stability properties of two-step Runge-Kutta methods for delay differential equations. The methods under consideration are Runge-Kutta methods that use the internal stages of two consecutive steps to define a polynomial approximation to the solution in the current step. Since all solution \( y(t)\) of the linear test equation \( y'(t)= a y(t) + b y(t- \tau)\) with \( \tau > 0 \) a constant delay and \(a\) and \(b\) complex coefficients such that \( |b |< \text{Re} (a)\) tends to zero as \( t \to + \infty\), a numerical method is said to be P-stable if this property holds for all numerical solutions with step sizes \( h = \tau / m \) for all positive integer \(m\). In this setting the authors prove that if the numerical method is A-stable (for linear ordinary differential equations \( y'(t) = a y(t)\) with \(\operatorname {Re}a<0\)) and satisfy some additional assumptions then it is also P-stable. Some examples of fourth order methods of this type are given and further it is shown that it is possible to extend a one step A-stable method to a two-step P-stable method.