Scholarlink

<p>Fusion zones (FZs) and heat-affected zones (HAZs) possess microstructure and mechanical properties different from those of the base material. Due to the small size of FZs and HAZs, directly obtaining the mechanical properties using miniature tensile specimens is difficult. In this study, we developed a Bayesian approach integrating experimental tests and finite element modeling (FEM) to predict the mechanical properties of the FZ and HAZ of dual-phase high-strength steels (DP980) after laser welding. A three-dimensional finite element model is created to simulate the indentation test, establishing the correlation between the mechanical properties (stress-strain curves) and microhardness. Then, a Kriging surrogate model is used to approximate the FEM indentation model, while the Monte Carlo Markov Chains (MCMC) are used to represent the correlations between stress-strain curves and microhardness. Later, a three-dimensional FEM, including the actual microstructure of the welding bead, is developed to simulate the tensile test for laser welding blanks (LWBs). Finally, the individual zone properties are determined based on the experimentally obtained microhardness and fracture location. The obtained zone properties accurately predict the tensile failure location, Erichsen cup height, and crack initiation in the Erichsen cupping test. The current inverse method needs fewer experimental tests and provides accurate predictions. Meanwhile, the new approach helps optimize laser welding process parameters and improve the LWB strength.</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

Predicting the Individual Heat-Affected Zone Properties of Laser Welded DP980 Using the Bayesian Approach

  • 期刊论文

  • 发表日期：

    2025年9月2日

查看全文
  • Yonghao Zhao,
  • Kaiyuan Chen,
  • Yu Zhang,
  • Guang Cheng

摘要

Fusion zones (FZs) and heat-affected zones (HAZs) possess microstructure and mechanical properties different from those of the base material. Due to the small size of FZs and HAZs, directly obtaining the mechanical properties using miniature tensile specimens is difficult. In this study, we developed a Bayesian approach integrating experimental tests and finite element modeling (FEM) to predict the mechanical properties of the FZ and HAZ of dual-phase high-strength steels (DP980) after laser welding. A three-dimensional finite element model is created to simulate the indentation test, establishing the correlation between the mechanical properties (stress-strain curves) and microhardness. Then, a Kriging surrogate model is used to approximate the FEM indentation model, while the Monte Carlo Markov Chains (MCMC) are used to represent the correlations between stress-strain curves and microhardness. Later, a three-dimensional FEM, including the actual microstructure of the welding bead, is developed to simulate the tensile test for laser welding blanks (LWBs). Finally, the individual zone properties are determined based on the experimentally obtained microhardness and fracture location. The obtained zone properties accurately predict the tensile failure location, Erichsen cup height, and crack initiation in the Erichsen cupping test. The current inverse method needs fewer experimental tests and provides accurate predictions. Meanwhile, the new approach helps optimize laser welding process parameters and improve the LWB strength.