29 Optimization of an aeronautical composite structure through a parallel multilevel approach Leonardo D. Chiwiacowsky 1 , Paolo Gasbarri 2 , Haroldo F. C. Velho 3 and Arthur T. G´omez 1 1 Universidade do Vale do Rio dos Sinos, S˜ao Leopolodo, Brazil 2 University of Rome “La Sapienza”, Rome, Italy 3 Instituto Nacional de Pesquisas Espaciais, S˜ao Jos´ e dos Campos, Brazil Summary. Optimal design of complex engineering systems, such as aircraft composite structures, can often be accomplished only by decomposition techniques. In general, it is characterized by a multidisciplinary task, usually involving multiple objectives, with many design variables and several constraints associated to each one of the disciplines taken into account. One of these techniques is represented by the multilevel approach, where the complex problem of multidisciplinary optimization is solved based on the idea “divide and conquer”. The original problem is split into several smaller subproblems, making the new conguration inherently suited to parallel and distributed computing. In this paper the optimal design of a composite wing-box is addressed by using a parallel two-level scheme, where a stochastic method is used to solve each second-level problem. At this level, since the changes in the components of one of the sub-problems have no influence on the solution of the others, they can be solved simultaneously. 1 INTRODUCTION A wing composite structure is composed of a large number of panels, which have to be designed simultaneously to obtain optimum structural design. Usually the design of each panel requires a large number of design variables to describe its geometry, ply composition and laminate stacking sequences. By virtues of these considerations, designing all the panels simultaneously constitutes an extremely large optimization problem, requiring also a very detailed structural model of the entire wing, that