Search

Abstract This paper presents a multi-level aeroelastic tailoring framework for the optimisation of composite aircraft wings. The framework is capable of structural sizing and produces detailed composite ply configurations through robust and reliability-based design optimisation, and is demonstrated on a representative regional jet airliner finite element wing box model. The optimisation procedure is divided into two levels. The first level optimises the wing structure for minimum weight subject to multiple constraints including strain, buckling, aeroelastic stability and gust response. These first level solutions are then fed into the second level to be further optimised for robustness or reliability by considering uncertainties in material properties at ply level. Both the principles of robust and reliability-based design optimisation can also be used in combination to ensure a balance between the robustness and reliability of the structural performance. In order to keep computations to an acceptable cost, the second level optimisation employs the Polynomial Chaos Expansion method to approximate the effect of probabilistic uncertainty on structural performance. In comparison to the original benchmark wing, the framework produces an overall weight reduction of 32.1%, despite a 1.5% increase from the first to the second level optimisation that accounts for stochastic design variations. [ABSTRACT FROM AUTHOR]

An approach is presented for modelling the aeroelastic stability of composite laminate wings with ply orientations subject to uncertainty. An aeroelastic model is constructed using the Rayleigh–Ritz technique coupled with modified strip theory aerodynamics. Lamination parameters are used as inputs to a Polynomial Chaos Expansion (PCE), enabling efficient propagation of the uncertainty through the aeroelastic model for a composite laminate with any number of plies. The Rosenblatt transformation is used to adapt the lamination parameter distributions for use with the PCE. A modified approach for modelling the uncertain aeroelastic response near the boundaries of different regimes of behaviour is also presented. Two case studies demonstrate application of the techniques; the first applies a simple PCE to a number of example balanced and symmetric laminates, the second applies the modified approach in a parametric investigation of the effects of bend–twist coupling on response mechanism. The proposed techniques offer at least an order of magnitude reduction in computation time compared to baseline Monte Carlo Simulation for all of the cases investigated. [ABSTRACT FROM AUTHOR]