Optimization research of stiffened shells based on kriging model and explicit FEM

• Published in: 2017 8th International Conference on Mechanical and Aerospace Engineering (ICMAE) pp. 305 - 308
• Main Authors: Wu Yongliang, Chen Jianping, Wei Wei, Wang Bin
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2017
• Subjects: Algorithm design and analysis; explicit FEM; Finite element analysis; Force; kriging surrogate model; Load modeling; Loading; Numerical models; Optimization; stiffened shells
• ISBN: 1538633051; 9781538633052
• DOI: 10.1109/ICMAE.2017.8038661

A novel parametric model of the stiifened shell is developed by Python language with Abaqus in this paper. As an effective analysis method in the optimal design, explicit finite element method (FEM) is used to design and optimize the skin thickness and stiffener size of Stiffened Shells. The optimization objectives and constraints are mainly the minimum mass subject to the structural performance and the high structural performance subject to the mass. In order to avoid the enormous computational cost of complex simulations in design optimization process, a surrogate kriging model is built based on the experimental design. The surrogate model method and the multi-island genetic algorithm are used to achieve a preliminary design based on the approximate model, and the sequential linear programming algorithm is employed to optimize continuous variables. So the computing time of optimization analysis is reduced greatly. The simulation results show that a reasonable and useful design can be realized with the optimization methods.