Simulation and Optimization for Hot Stamping Process of Rear Windshield Lower Crossbeam of Aluminum Alloy Automobiles

Authors

  • Huijun ZHAO Hangzhou Dianzi University https://orcid.org/0000-0002-2737-2416
  • Jiarui XU Hangzhou Dianzi University
  • Jianghua HUANG Zhejiang University of Science and Technology
  • Yang XIAO Hangzhou Dianzi University
  • Haibing FU Zhejiang Zhaofeng Mechanical and Electronical CO., LTD.
  • Binghe JIANG Zhejiang University of Science and Technology

DOI:

https://doi.org/10.5755/j02.mech.40111

Keywords:

aluminum alloy, hot stamping, machine learning, process optimization, mechanical simulation

Abstract

The simulation of the hot stamping process for the rear windshield lower crossbeam in aluminum alloy automobiles was carried out. Contrasted with the one-step forming process, the maximum thickness increase rate was lowered by 6.1 % and the maximum thinning rate was decreased by 24.8 % using the two-step forming process. Under identical process parameters, the minimum thickness of two-step formed parts was greater than that of one-step formed parts. The forming experiment was conducted by employing the self-developed modular hot stamping die.A gradient boosting regression tree machine learning model for the maximum thickening rate and the maximum reduction rate of the parts was established, and the optimal process parameters matching for the hot stamping of the rear windshield lower crossbeam were obtained based on the NSGA-II multi-objective optimization algorithm, namely, forming temperature 570 °C, friction coefficient 0.15, stamping speed 450 mm/s and die clearance 1.05 t.

Downloads

Published

2026-03-27

Issue

Section

MECHANICAL TECHNOLOGIES