A Novel Reduced‐order Modeling Method for Nonlinear Buckling Analysis and Optimization of Geometrically Imperfect Cylinders

Ke Liang,Peng Hao,Bo Wang,Qin Sun
DOI: https://doi.org/10.1002/nme.6585
IF: 3.021
2020-01-01
International Journal for Numerical Methods in Engineering
Abstract:A novel reduced‐order modeling method with two‐step strategy is proposed for nonlinear buckling analysis of axially compressed thin‐walled cylinder. The nonlinear response analysis up until the buckling point is achieved quickly by solving the small‐scale reduced‐order model, accounting for the geometrically nonlinear behavior. The buckling point is determined accurately using an efficient buckling detection algorithm, where the stiffness information of the reduced‐order model is extrapolated based on an eigenvalue analysis until a singular tangent stiffness is obtained. Then, the nonlinear buckling fiber angle optimization scheme of laminated composite cylinders is constructed in the MATLAB's MultiStart scheme with gradient‐based algorithms. The proposed two‐step reduced‐order modeling strategy is adopted to calculate the nonlinear buckling objection function at a much lower cost. Numerical results for axially compressed cylinders with various geometric imperfection fields demonstrate the good performance of the proposed strategy in both nonlinear buckling analyses and lamination optimizations.
What problem does this paper attempt to address?