Scholarlink

<p>Thin-walled parts refer to lightweight structural parts comprised of thin plates and stiffeners. During the machining process of thin-walled parts, machining distortion often occurs due to uncertain factors such as varying stiffness, cutting force, cutting temperature, and&#xa0;residual stress. In the paper, a new approach for quantifying the influence of initial residual stress on machining distortion uncertainty is proposed by combining the gradient boosting regression tree algorithm and finite element simulation. Firstly, finite element simulation is used to model the milling process of the workpiece and obtain machining distortion data. These data are validated using small deflection theory of thin plates. Next, the machining distortion simulation data is expanded using a gradient boosting regression tree to learn the coupling influences between material removal rate, inherent uncertainty in initial residual stresses, and machining distortion uncertainty. Finally, based on the predictive data from the gradient boosting regression tree, the machining distortion uncertainty inference is conducted, indirectly quantifying the influence of redistribution of initial residual stresses and the uneven distribution of these stresses on machining distortion uncertainty. Among them, the influential weight of the initial residual stress redistribution on the machining distortion uncertainty is 0.903, and the influential weight of uneven distribution of these stresses on the machining distortion uncertainty is 0.097. Taking the web structure and T-shaped thin-walled parts as examples, the measures to reduce the influence of initial residual stress redistribution on the machining distortion uncertainty are investigated. It is found that the machining distortion uncertainty caused by the strategy of spiral milling from inwards to outwards is the smallest, only 0.012. Furthermore, in addition to strictly controlling the finishing allowance, the finishing allowance should be symmetrically distributed near the neutral layer as far as possible.</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

Investigation on the influence of initial residual stress on machining distortion uncertainty

  • 期刊论文

  • 发表日期：

    2025年5月30日

查看全文
  • Zhaoze Sun,
  • Xiaoyue Li,
  • Hao Qi,
  • Mengqi Chen

摘要

Thin-walled parts refer to lightweight structural parts comprised of thin plates and stiffeners. During the machining process of thin-walled parts, machining distortion often occurs due to uncertain factors such as varying stiffness, cutting force, cutting temperature, and residual stress. In the paper, a new approach for quantifying the influence of initial residual stress on machining distortion uncertainty is proposed by combining the gradient boosting regression tree algorithm and finite element simulation. Firstly, finite element simulation is used to model the milling process of the workpiece and obtain machining distortion data. These data are validated using small deflection theory of thin plates. Next, the machining distortion simulation data is expanded using a gradient boosting regression tree to learn the coupling influences between material removal rate, inherent uncertainty in initial residual stresses, and machining distortion uncertainty. Finally, based on the predictive data from the gradient boosting regression tree, the machining distortion uncertainty inference is conducted, indirectly quantifying the influence of redistribution of initial residual stresses and the uneven distribution of these stresses on machining distortion uncertainty. Among them, the influential weight of the initial residual stress redistribution on the machining distortion uncertainty is 0.903, and the influential weight of uneven distribution of these stresses on the machining distortion uncertainty is 0.097. Taking the web structure and T-shaped thin-walled parts as examples, the measures to reduce the influence of initial residual stress redistribution on the machining distortion uncertainty are investigated. It is found that the machining distortion uncertainty caused by the strategy of spiral milling from inwards to outwards is the smallest, only 0.012. Furthermore, in addition to strictly controlling the finishing allowance, the finishing allowance should be symmetrically distributed near the neutral layer as far as possible.