Crashworthiness analysis and structural optimization of shrink tube under interference condition

Autor: Ping Xu, Yuhui Yang, Chengxing Yang, Shuguang Yao, Jie Xing, Fan Zou
Jazyk: angličtina
Rok vydání: 2023
Předmět:
Zdroj: Engineering Science and Technology, an International Journal, Vol 46, Iss , Pp 101504- (2023)
Druh dokumentu: article
ISSN: 2215-0986
DOI: 10.1016/j.jestch.2023.101504
Popis: This paper aims to study the crashworthiness of shrink tubes for rail vehicles under axial impact loading. A theoretical model of shrink tube for platform force prediction under interference fit conditions was proposed. The trolley crash test was used to experimentally study the structural energy absorption and characteristics during collision, and a corresponding finite element model was constructed and validated. The finite element model validated the accuracy of the theoretical model. The influence of structural parameters on the crashworthiness of shrink tube under interference conditions was then studied. Structural design variables, including the magnitude of interference (int_pre), thickness (T), long right-angle edge (tri_x) and short right-angle edge (tri_z) were sampled using a full factorial and a Latin hypercube design method. Based on the above samples, the moving least squares method (MLSM) was used to construct an approximate model of peak force (Fmax), specific energy absorption (SEA), and platform force (Fplt). The main effects analysis illustrated that the tri_z had the most significant effects on the peak force, specific energy absorption and platform force. To minimize Fmax and maximize SEA, the global response surface method (GRSM) was employed for parameter optimization. Finally, the optimized configurations were obtained with int_pre=0.067mm, T=10.003mm, tri_x=21.485mm, and tri_z=6.101mm. Compared to the initial design, the SEA value was increased by 16.41 %, while the Fmax value was decreased by 2.21%. The error of platform force derived from the optimal solution and theoretical prediction was 3.62%, which verified that the theoretical prediction model was credible within an acceptable range.
Databáze: Directory of Open Access Journals