Progressive Optimization of Complex Shells with Cutouts Using a Smart Design Domain Method
Peng Hao,Yu Wang,Zhangming Wu,Xuanxiu Liu,Bo Wang,Wei Huang
DOI: https://doi.org/10.1016/j.cma.2019.112814
IF: 6.588
2020-01-01
Computer Methods in Applied Mechanics and Engineering
Abstract:Thin-walled shells with cutouts are widely used as primary structures in the aerospace field. The mechanical analysis of shell structures using the Finite Element Analysis (FEA) method often needs very fine meshes to achieve a high-fidelity simulation. The advent of isogeometric analysis (IGA) method provides an alternative, yet efficient and accurate means to model complex shell structures with cutouts. In this paper, the trimmed surface analysis (TSA) method is applied in IGA to perform the trimmed elements integration of linear buckling analysis of shell structures. In the shape optimization process, the analytical formulae for the design sensitivities of IGA based shell buckling model are derived, and the sensitivity propagation from the design model to the analysis model is computed using a h-refinement method. Moreover, a novel smart design domain (SDD) method is proposed and implemented to reduce the number of design variables and further enhance the shape optimization efficiency, substantially. SDD can provide useful reference information to guide the selection of control points as design variables. Later, the comparison of optimization examples approve that the SDD method can provide a very efficient means to perform the local shape optimization for the complex shells. Finally, a multi-level progressive optimization framework based on the SDD method is developed, in which the design model and the analysis model are separated and implemented independently. In doing so, the design model is able to provide enough parameterization for optimization, and, in the meantime, the analysis model ensures efficiency and accuracy for simulation. Finally, the numerical examples on the buckling analysis and optimization of complex shells with cutouts demonstrate the greatly improved capacity and efficiency of the design framework proposed in this paper.