<h2>AAS 95-420</h2><h2>Runge-Kutta Integration Methods for Trajectory Propagation Revisited</h2><h4>G. J. Der, TRW , Redondo Beach, CA</h4><h2> Abstract </h2>Runge-Kutta integration methods are simple to use, stable and self-starting.  This paper revisits Runge-Kutta integration methods for near Earth ballistic trajectories influenced by atmospheric drag and acceleration due to gravitational harmonics.  Accuracy and computer timing results are presented for several Runge-Kutta integration methods.  For trajectory propagation, it is concluded that:  (1) There is no atmospheric density model uniformly good at all altitudes.  (2) There is no best classical Runge-Kutta integration method uniformly good for all applications.  (3) A higher order classical Runge-Kutta integration method may be 10 times more computationally efficient than a lower order classical Runge-Kutta integration method.<br><br>