Integration of computational fluid dynamics and control system for a missile-shaped body

© 2012 Jonathan Brian Da CostaThe simulation of missile systems plays an important role in the development of technological improvements for defence forces worldwide due to the considerable cost advantages. Within this field, aerodynamics and control continue to be main focal areas that can greatly enhance mission critical parameters such as range and manoeuvrability. Computational Fluid Dynamics (CFD) software has evolved over several decades as an engineering tool for performing flow analysis around complex aerodynamic bodies. From CFD data, an analysis can be performed to characterise the forces and moments that a bluff body experiences in a uniform flow. The data gathered from a CFD simulation can be used as the basis for a low order aerodynamic model, and subsequently used in control design. As missile path tracking represents a constrained control problem, Linear Quadratic Regulation (LQR) presents a challenging approach to optimal control system design. This thesis examines the steady state aerodynamic characteristics of a missile-shaped bluff body in a supersonic free stream and uses these characteristics as the basis for control system design. Confidence in the Reynolds Averaged Navier Stokes (RANS) simulation fidelity is developed systematically by using classical aerodynamic examples, a sphere and an ogive-cylinder, before adding fins for the full missile model. Numerical data from the simulated bluff bodies is found to compare well with published experimental data. The aerodynamic coefficients are supplied to a control-oriented, nonlinear, Six Degree of Freedom (6DoF) model. Derivation of the control system addresses the inherent computational complexity of the nonlinear dynamics and kinematics through linearisation of the 6DoF model using a Linear Time Varying (LTV) approach, coupled with full state feedback. The control system is subsequently tested by perturbing the missile mid-flight trajectory and measuring the resultant deviation from a nominal flight path. Under the proposed approach, the control system counteracts positional and angular error within an appropriate time frame, even in the presence of actuator saturation