|dc.description.abstract||The use of computer programs driven by genetic algorithms (GA’s) has become an increasingly popular method of optimizing engineering designs. This thesis focuses on the modeling and optimization of liquid rocket engine propelled missiles with an emphasis on some recent upgrades to an existing suite of codes. The program is comprised of a series of legacy codes which simulate the performance of liquid rockets and are controlled by a GA. This program is designed so that it can be used to reverse engineer missile designs for which some limited initial data is known. It can also be used to optimize liquid propelled missiles in the more traditional design mode.
Several upgrades were made to the existing code to expand its capabilities and add to the robustness of the program. A new version of the code has been developed in which aerodynamic prediction duties are handled by Missile Datcom instead of the Aerodsn routine used in previous versions. Plots from runs using the two different aerodynamic codes are presented and their differences discussed. Furthermore, the program now has the ability to handle a varying specific impulse (Isp), and a jet vane control system model has been added. The previous version of the liquid missile GA code assumed that Isp was constant for a given fuel type. The modification described in this thesis gives the GA the ability to choose an equivalence ratio which determines the Isp. In addition to the aerodynamic control system already in place, the ability to simulate a missile controlled by jet vanes has been added to the program. The new control system and accompanying optimization results are examined in detail.||en_US