Scholarlink

<p>Hot stamping of automotive structural parts involves strong thermal–mechanical coupling and multiple interacting process parameters, making rapid and accurate process optimization difficult when only finite element simulations are used. In this study, a BP–Kriging surrogate model combined with NSGA-II and entropy-weight TOPSIS was developed to predict and optimize the hot stamping process parameters of a rocker panel. The proposed model improved the prediction accuracy for maximum wrinkling, maximum thinning, and average hardness compared with a conventional Kriging model, with R² increased by 6.8%–12.7%, RMSE reduced by 15.4%–27.0%, and MAE reduced by 11.5%–27.5%. The optimal process parameters were determined as follows: sheet temperature of 922&#xa0;°C, stamping speed of 175&#xa0;mm·s⁻¹, binder force of 120 kN, pad force of 126 kN, and friction coefficient of 0.19. Under these conditions, the maximum thinning and maximum wrinkling of the formed part were reduced by 17.8% and 22.3%, respectively, while the average hardness reached 464 HV. Experimental verification further confirmed that the optimized process effectively suppressed cracking and wrinkling. The proposed framework provides an efficient approach for process-window design and rapid parameter optimization in hot stamping.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction and optimisation of hot stamping process parameters for the rocker panel based on the BP-Kriging model

  • 期刊论文

  • 发表日期：

    2026年7月20日

查看全文
  • Ruijia Qu,
  • Zhiqiang Zhang,
  • Mingwen Ren,
  • Hongjie Jia,
  • Weidong Zhu

摘要

Hot stamping of automotive structural parts involves strong thermal–mechanical coupling and multiple interacting process parameters, making rapid and accurate process optimization difficult when only finite element simulations are used. In this study, a BP–Kriging surrogate model combined with NSGA-II and entropy-weight TOPSIS was developed to predict and optimize the hot stamping process parameters of a rocker panel. The proposed model improved the prediction accuracy for maximum wrinkling, maximum thinning, and average hardness compared with a conventional Kriging model, with R² increased by 6.8%–12.7%, RMSE reduced by 15.4%–27.0%, and MAE reduced by 11.5%–27.5%. The optimal process parameters were determined as follows: sheet temperature of 922 °C, stamping speed of 175 mm·s⁻¹, binder force of 120 kN, pad force of 126 kN, and friction coefficient of 0.19. Under these conditions, the maximum thinning and maximum wrinkling of the formed part were reduced by 17.8% and 22.3%, respectively, while the average hardness reached 464 HV. Experimental verification further confirmed that the optimized process effectively suppressed cracking and wrinkling. The proposed framework provides an efficient approach for process-window design and rapid parameter optimization in hot stamping.