Abstract
The Steger-Warming flux vector splitting is used for the solution of two-dimensional Euler equations. An implicit finite difference scheme is used. A computer code is developed and tested for subsonic, transonic, and supersonic flow regimes. The method is then applied to the flow through a convergent-divergent nozzle and a divergent nozzle, both with strong shocks. Further factorization has been done to the formulated finite difference equations. As a result of the new factorization, the CPU time to reach the steady state solution decreased significantly to about 65% without altering the accuracy. Comprehensive comparisons with the MacCormack method are presented.