Multi-fidelity optimization of laminated conical shells for buckling

• Published in: Structural and multidisciplinary optimization Vol. 30; no. 2; pp. 128 - 141
• Main Authors: Goldfeld, Y., Vervenne, K., Arbocz, J., van Keulen, F.
• Format: Journal Article
• Language: English
• Published: Heidelberg Springer Nature B.V 01.08.2005
• Subjects: Accuracy; Buckling; Computational efficiency; Conical shells; Dependence; Filament winding; Load; Material properties; Mathematical analysis; Minimum weight; Optimization; Ply orientation; Response surface methodology; Stiffness matrix
• ISSN: 1615-147X; 1615-1488
• DOI: 10.1007/s00158-004-0506-9

Optimum laminate configuration for minimum weight of filament-wound laminated conical shells is investigated subject to a buckling load constraint. In the case of a composite laminated conical shell, due to the manufacturing process, the thickness and the ply orientation are functions of the shell coordinates, which ultimately results in coordinate dependence of the stiffness matrices (A,B,D). These effects influence both the buckling load and the weight of the structure and complicate the optimization problem considerably. High computational cost is involved in calculating the buckling load by means of a high-fidelity analysis, e.g. using the computer code STAGS-A. In order to simplify the optimization procedure, a low-fidelity model based on the assumption of constant material properties throughout the shell is adopted, and buckling loads are calculated by means of a low-fidelity analysis, e.g. using the computer code BOCS. This work proposes combining the high-fidelity analysis model (based on exact material properties) with the low-fidelity model (based on nominal material properties) by using correction response surfaces, which approximate the discrepancy between buckling loads determined from different fidelity analyses. The results indicate that the proposed multi-fidelity approaches using correction response surfaces can be used to improve the computational efficiency of structural optimization problems.