Buckling optimization of composite rectangular plates by sequential permutation search with bending-twisting correction

Buckling problem of composite plates is of importance since it is commonly encountered in thin-walled structural components in engineering. Buckling resistance design of composite plates involves the design of laminate stacking sequence exhibiting high dimensional discrete variables owing to the combination and permutation of ply orientations of layers. In this work, a sequential permutation search (SPS) optimization algorithm is presented for the stacking sequence design of composite rectangular plates accounting for the inherent physical-mechanical behavior of composite laminates. Owing to that the governing equation for buckling of composite rectangular plates is a linear homogeneous equation on bending stiffness parameters, the convexity of flexural lamination parameters and sensitivity of bending stiffness are utilized to construct sensitive detection techniques and linear search procedures. Moreover, to increase the robustness of SPS, a sign optimization algorithm (SOA) is coupled in SPS to regulate bending-twisting coupling effects. The SPS algorithm is applied to the buckling resistance design of composite rectangular plates exhibiting various boundaries under axial compression, where the Rayleigh-Ritz method is developed to perform the buckling analysis. Optimal results of SPS are compared against layerwise optimization approach (LOA) and genetic algorithm (GA), illustrating its reliability and efficiency. (c) 2021 Elsevier Inc. All rights reserved.

Automated summary

This paper introduces a sequential permutation search (SPS) optimization algorithm for the stacking sequence design of composite rectangular plates, utilizing the convexity of flexural lamination parameters and sensitivity of bending stiffness to construct sensitive detection techniques and linear search procedures. Coupled with a sign optimization algorithm (SOA) to regulate bending-twisting coupling effects, the SPS algorithm is shown to be reliable and efficient through comparisons with other methods.