Scholarlink

<p>The fabrication quality of film cooling holes (FCHs) in nickel-based single-crystal superalloys is crucial for ensuring the performance and reliability of aero-engines. However, traditional optimization methods often rely on extensive experiments and fail to achieve optimal results. This study proposes an integrated optimization framework that combines experimental analysis, machine learning, and multi-objective optimization to address this issue. Single-factor and orthogonal experiments were first conducted to identify the influence and importance of key laser processing parameters. A generative adversarial network (GAN) was then used to expand the dataset. Subsequently, a backpropagation neural network (BP) optimized by the whale optimization algorithm (WOA) was developed to model the nonlinear relationships between process parameters and output quality metrics, including hole taper and material removal rate. Finally, the WOA-BP model was integrated with the NSGA-II algorithm to determine a Pareto-optimal set of parameter combinations. The optimized results achieved prediction errors below 2%, with improvements of 10% in hole taper and 6% in material removal rate, demonstrating the effectiveness of the proposed approach.</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

Parameter Influence and Optimization Strategy for Femtosecond Laser Drilling of Nickel-Based Single-Crystal Superalloys

  • 期刊论文

  • 发表日期：

    2025年9月22日

查看全文
  • Yuanmin Tu,
  • Zhenwei Li,
  • Zhixun Wen,
  • Pengfei He,
  • Ying Dai

摘要

The fabrication quality of film cooling holes (FCHs) in nickel-based single-crystal superalloys is crucial for ensuring the performance and reliability of aero-engines. However, traditional optimization methods often rely on extensive experiments and fail to achieve optimal results. This study proposes an integrated optimization framework that combines experimental analysis, machine learning, and multi-objective optimization to address this issue. Single-factor and orthogonal experiments were first conducted to identify the influence and importance of key laser processing parameters. A generative adversarial network (GAN) was then used to expand the dataset. Subsequently, a backpropagation neural network (BP) optimized by the whale optimization algorithm (WOA) was developed to model the nonlinear relationships between process parameters and output quality metrics, including hole taper and material removal rate. Finally, the WOA-BP model was integrated with the NSGA-II algorithm to determine a Pareto-optimal set of parameter combinations. The optimized results achieved prediction errors below 2%, with improvements of 10% in hole taper and 6% in material removal rate, demonstrating the effectiveness of the proposed approach.