Optimization of the Conceptual Design of a Multistage Rocket Launcher
Pedro Orgeira-Crespo,
Guillermo Rey,
Carlos Ulloa,
Uxia Garcia-Luis,
Pablo Rouco,
Fernando Aguado-Agelet
Affiliations
Pedro Orgeira-Crespo
Department of Mechanical Engineering, Heat Engines and Machines and Fluids, Aerospace Engineering School, University of Vigo, Campus Orense, 32004 Orense, Spain
Guillermo Rey
Department of Mechanical Engineering, Heat Engines and Machines and Fluids, Aerospace Engineering School, University of Vigo, Campus Orense, 32004 Orense, Spain
Carlos Ulloa
Department of Mechanical Engineering, Heat Engines and Machines and Fluids, Aerospace Engineering School, University of Vigo, Campus Orense, 32004 Orense, Spain
Uxia Garcia-Luis
Department of Mechanical Engineering, Heat Engines and Machines and Fluids, Aerospace Engineering School, University of Vigo, Campus Orense, 32004 Orense, Spain
Pablo Rouco
Department of Mechanical Engineering, Heat Engines and Machines and Fluids, Aerospace Engineering School, University of Vigo, Campus Orense, 32004 Orense, Spain
Fernando Aguado-Agelet
Telecommunication Engineering School, University of Vigo, 36310 Vigo, Spain
The design of a vehicle launch comprises many factors, including the optimization of the climb path and the distribution of the mass in stages. The optimization process has been addressed historically from different points of view, using proprietary software solutions to obtain an ideal mass distribution among stages. In this research, we propose software for the separate optimization of the trajectory of a launch rocket, maximizing the payload weight and the global design, while varying the power plant selection. The launch is mathematically modeled considering its propulsive, gravitational, and aerodynamical aspects. The ascent trajectory is optimized by discretizing the trajectory using structural and physical constraints, and the design accounts for the mass and power plant of each stage. The optimization algorithm is checked against various real rockets and other modeling algorithms, obtaining differences of up to 9%.
