Lay-Up Optimization of Composite Stiffened Panels using Linear Approximations in Lamination Space

A new two-step approach to optimize anisotropic composite stiffened panels is presented. At the first step, a representative element of the stiffened panel (superstiffener) is optimized using continuous optimization of lamination parameters under strength, buckling and practical design constraints. At the second step, a genetic algorithm is used to identify the actual superstiffener’s laminates. The fitness function in the genetic algorithm is formed by using a first order (linear) Taylor series of the design constraints, instead of the traditional squared differences between the optimum and actual lamination parameters (minimum squared distance). Results show that for the same thicknesses of the superstiffener’s laminates, the new designs had lower violation of the critical constraint than those obtained from minimum squared distanced. Consequently, laminates’ thicknesses could be reduced and thus mass savings are achieved. In addition, it is found that fitness based on constraint satisfaction drives the genetic algorithm in a different direction than the minimum distance criterion and produces designs that may not be close to the continuous design in the lamination parameter design space. Overall, this suggests that the minimum squared distance may not be the best objective to identify the optimal laminates’ stacking sequences.