Methods for Optimization of a Launch Vehicle for Pressure Fluctuation Levels and Axial Force

Abstract

A computational fluid dynamics (CFD) code has been combined with a Genetic Algorithm (GA) to perform a shape optimization study on a two dimensional axisymmetric model of a typical launch vehicle. The objective of this study was to demonstrate a methodology for reducing pressure fluctuations and the axial force coefficient for a launch vehicle throughout a typical ascent trajectory. Due to the high computational expense and difficulty of generating an adequate mesh autonomously, few CFD driven GA optimizations have been conducted. Some of the complexity of this process was alleviated by using a simple two dimensional axisymmetric geometry to model the vehicle.

The optimization process involved the GA selecting a set of geometric parameters that define the shape of the vehicle. A grid generator created a mesh based on these parameters and a CFD solver calculated the flow parameters. The grid generator is a FORTRAN routine written for this particular geometric shape. The FORTRAN code created a mesh file dependent only on the geometric variables chosen by the GA. The pressure fluctuation level and axial force coefficient are calculated by the flow parameters that are obtained from the CFD solution.

A pressure fluctuation level minimization study and axial force minimization study were conducted separately using the same CFD model. The results of each optimization study were compared to a baseline geometry having a very similar shape to the Ares I Crew Launch Vehicle. The results of the pressure fluctuation study yielded a reduction in the average RMS pressure fluctuation level throughout the ascent trajectory. The average RMS fluctuating pressure level was reduced by approximately 17.5% compared to the baseline geometry; however the optimized geometry would not be favorable as a practical design for a launch vehicle shape. While the resulting optimized geometry for the pressure fluctuation study is not an ideal design, the methodology for reducing pressure fluctuations using a GA combined with CFD is shown. The axial force minimization study yielded a reduction in the axial force coefficient of approximately 56%. The resulting shape from the axial force minimized solution was found to resemble that of a blunted ogive, as expected.