Subdomain hybrid cellular automata method for material optimization of thin-walled frame structure under transverse impact. (April 2023)

• Record Type: Journal Article
• Authors: Duan, Libin; Xu, Zheng; Xu, Wei; Zhang, Xuerong; Du, Zhanpeng; Liu, Xing; Jiang, Haobin
• Is Part Of: International journal of impact engineering. Volume 174(2023)
• Publication Date: 2023-04
• Subjects: Subdomain cellular automata -- Material optimization -- Thin-walled frame structure -- Nonlinear dynamic response optimization -- Car body frame; Impact -- Periodicals; Shock (Mechanics) -- Periodicals; Impact -- Périodiques; Choc (Mécanique) -- Périodiques; Impact; Shock (Mechanics); Periodicals; 620.1125
• DOI: 10.1016/j.ijimpeng.2023.104524
• Languages: English
• ISSNs: 0734-743X

Abstract

Abstract: In this paper, a subdomain hybrid cellular automata method is proposed to carry out the material optimization of thin-walled frame structure under transverse impact with minimum material cost, which is a typical optimal solution searching problem including many discrete variables. A nominal flow stress is introduced to convert the discrete design variables (cell materials) to be continuous for the update of cell materials. The global optimal searching ability of the proposed method is improved by the material update rule based on the step internal energy density target, which is adaptively defined during the material optimization to avoid the optimal solution jumping into a local one. PID control is employed to improve the robustness of the proposed method. The subdomain cellular automata model of a car body frame under transverse impact is constructed to separately perform the material and cost optimization by the proposed method and a parallel efficient global optimization method. The total material cost converges to CNY 1051 within 271 iterations using 32 CPUs by the parallel method and CNY 1101 within 8 iterations using 8 CPUs by the proposed method. The material optimization efficiency is significantly improved by the proposed method with good global searching ability. Highlights: New material optimization method of nonlinear dynamic problems with discrete variables. Significant optimization efficiency improvement with the proposed method. Global optimization improved by material update rule with step internal energy density. No gradient required, so especially suitable for nonlinear problems with discrete variables.