Akademik Veri Yönetim Sistemi
Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Havacılık ve Uzay Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2015
Öğrenci: ENGİN USTA
Danışman: İSMAİL HAKKI TUNCER
Özet:Missiles with very low aspect ratio fins (also called strakes) are generally difficult to model by using engineering level aerodynamic methods and linear theories break due to complex nature of the flow adjacent to the mutual interference of the body and fin. Since fast prediction methods cannot accurately model vortex development along and downstream of very low aspect ratio fins, they have limited success for application to missiles with strake fins. Moreover, there is limited number of CFD studies for investigating aerodynamics of very low aspect ratio fins. In this thesis, a generic test model is selected and CFD analyses are performed to estimate the normal force and center of pressure of the strake unit with varying angle of attack at supersonic Mach numbers for two different roll orientations. The CFD results are compared with fast prediction methods (Missile DATCOM and White’s method) and wind tunnel data. By the comparative validation study, CFD analyses are proven to be reliable and robust for application to missiles with very low aspect ratio fins. In addition, a limited goal CFD based optimization study to design strake planform geometry is performed for alleviating the cross coupling effect on the pitching moment from sideslip angle while keeping static margin in a constant range. The geometrical parameters varying during the optimization problem consist of the leading edge location, chord length and span length of the strake planform. Since the configuration is monotype consisting of a body with strake and tail, automatic meshing scheme is developed and the whole processes from grid generation to CFD solutions are fully automated. For optimization problem, firstly a gradient based method is tested and then a stochastic optimization method is employed to guarantee global optimization instead of finding local minimum. Particle Swarm Optimization Method (PSO) is selected as a suitable population based stochastic optimization method for this study. The gradient based optimization method namely the steepest descent with line search is also applied for the same problem to investigate the applicability of gradient based optimization methods. It was observed that at the end of the PSO iterations, a small chord, short span strake comes out. However, it was also observed that the gradient solution ends with a short span, long chord strake geometry. Apparently, the gradient based optimization method is trapped in the local minimum. By the PSO optimization study, the cross coupling on the pitching moment from sideslip is decreased to one third of the initial value from baseline to optimum configuration while static margin is kept in a particular range.