International Journal of Machine Tools & Manufacture最新文献_第2页

AI-powered semiconductor wafer fabrication: A manufacturing paradigm shift 人工智能驱动的半导体晶圆制造：制造模式的转变

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-11-07 DOI: 10.1016/j.ijmachtools.2025.104345

Guanwei He , Qingqing Huang , Xinhao Li , Libo Zhou , Miao Yang , Yadan Luo , Marian Wiercigroch , Han Huang

Artificial intelligence (AI) is reshaping the landscape of modern manufacturing by enabling intelligent, adaptive, and increasingly autonomous production systems. In high-end manufacturing processes, persistent challenges such as complex machine-tool dynamics, variability in equipment performance, and environmental instabilities hinder consistent quality and high efficiency. These challenges are further amplified in advanced manufacturing domains, particularly in semiconductor wafer fabrication, which demands ultra-high precision to be attained. AI offers powerful solutions to these challenges through its capability in real-time decision making, data driving modelling, and predictive analysis of manufacturing processes. By learning from vast datasets generated during manufacturing, AI systems can identify working patterns, optimize process parameters, detect anomalies, and guide autonomous control strategies. These advantages position AI as a key enabler for enhancing process understanding, ensuring quality assurance, and accelerating innovation in fabrication technologies. Semiconductor wafer fabrication is the cornerstone of modern electronics manufacturing, serving as the foundation process for integrated circuits production. This article reviews the transformative role of AI in wafer fabrication, highlighting its principles, applications, and prospects. Key AI technologies such as deep learning, reinforcement learning, and generative algorithms that have been widely adopted in wafer fabrication are surveyed. Special attention is given to their deployment across critical fabrication procedures including crystal growth, ingot slicing and ultraprecision machining. The review further presents state-of-the-art research and industrial implementations of AI in these domains, showcasing how intelligent models can be employed in quality prediction, regime classification, process optimization, and system control. Finally, emerging trends and future perspectives aimed at fostering more practical, deeply integrated, and robust AI applications across the manufacturing ecosystem are discussed, paving the way toward fully autonomous and self-optimizing future manufacturing processes.

人工智能（AI）通过实现智能、自适应和日益自主的生产系统，正在重塑现代制造业的格局。在高端制造过程中，持续的挑战，如复杂的机床动力学、设备性能的可变性和环境的不稳定性，阻碍了始终如一的质量和高效率。这些挑战在先进制造领域进一步放大，特别是在半导体晶圆制造领域，这需要达到超高的精度。人工智能通过其实时决策、数据驱动建模和制造过程预测分析的能力，为这些挑战提供了强大的解决方案。通过学习制造过程中产生的大量数据集，人工智能系统可以识别工作模式、优化工艺参数、检测异常并指导自主控制策略。这些优势使人工智能成为增强工艺理解、确保质量保证和加速制造技术创新的关键推动者。半导体晶圆制造是现代电子制造业的基石，是集成电路生产的基础工艺。本文综述了人工智能在晶圆制造中的变革作用，重点介绍了其原理，应用和前景。综述了在晶圆制造中广泛采用的关键人工智能技术，如深度学习、强化学习和生成算法。特别关注它们在关键制造过程中的部署，包括晶体生长，铸锭切片和超精密加工。本文进一步介绍了人工智能在这些领域的最新研究和工业实现，展示了智能模型如何应用于质量预测、状态分类、过程优化和系统控制。最后，讨论了旨在促进整个制造生态系统中更实用、深度集成和强大的人工智能应用的新兴趋势和未来前景，为实现完全自主和自我优化的未来制造流程铺平了道路。

{"title":"AI-powered semiconductor wafer fabrication: A manufacturing paradigm shift","authors":"Guanwei He , Qingqing Huang , Xinhao Li , Libo Zhou , Miao Yang , Yadan Luo , Marian Wiercigroch , Han Huang","doi":"10.1016/j.ijmachtools.2025.104345","DOIUrl":"10.1016/j.ijmachtools.2025.104345","url":null,"abstract":"<div><div>Artificial intelligence (AI) is reshaping the landscape of modern manufacturing by enabling intelligent, adaptive, and increasingly autonomous production systems. In high-end manufacturing processes, persistent challenges such as complex machine-tool dynamics, variability in equipment performance, and environmental instabilities hinder consistent quality and high efficiency. These challenges are further amplified in advanced manufacturing domains, particularly in semiconductor wafer fabrication, which demands ultra-high precision to be attained. AI offers powerful solutions to these challenges through its capability in real-time decision making, data driving modelling, and predictive analysis of manufacturing processes. By learning from vast datasets generated during manufacturing, AI systems can identify working patterns, optimize process parameters, detect anomalies, and guide autonomous control strategies. These advantages position AI as a key enabler for enhancing process understanding, ensuring quality assurance, and accelerating innovation in fabrication technologies. Semiconductor wafer fabrication is the cornerstone of modern electronics manufacturing, serving as the foundation process for integrated circuits production. This article reviews the transformative role of AI in wafer fabrication, highlighting its principles, applications, and prospects. Key AI technologies such as deep learning, reinforcement learning, and generative algorithms that have been widely adopted in wafer fabrication are surveyed. Special attention is given to their deployment across critical fabrication procedures including crystal growth, ingot slicing and ultraprecision machining. The review further presents state-of-the-art research and industrial implementations of AI in these domains, showcasing how intelligent models can be employed in quality prediction, regime classification, process optimization, and system control. Finally, emerging trends and future perspectives aimed at fostering more practical, deeply integrated, and robust AI applications across the manufacturing ecosystem are discussed, paving the way toward fully autonomous and self-optimizing future manufacturing processes.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"214 ","pages":"Article 104345"},"PeriodicalIF":18.8,"publicationDate":"2025-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145461562","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Machine learning applications in welding processes: Progresses and opportunities 机器学习在焊接工艺中的应用：进展与机遇

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-11-03 DOI: 10.1016/j.ijmachtools.2025.104344

Peihao Geng , Yujun Xia , Zhiqiao Dong , Boxuan Men , Bo Pan , Chenhui Shao , Yongbing Li , Jingjing Li

The increasing demand for intelligent and autonomous manufacturing has driven the integration of machine learning (ML) into modern welding processes. Owing to its ability to model nonlinear and cross-scale interactions and extract critical features from complex, high-dimensional data, ML is rapidly transforming the design, monitoring, and evaluation of welding processes. Based on this, the paper systematically reviews research progress in ML for four representative welding processes (arc, laser, resistance and friction stir welding) over the past decade. First, typical welding tasks are categorized into three domains: pre-weld design, in-process monitoring, and post-weld quality assessment. It then elaborates on the types of welding data used and their input-output relationships across different tasks and analyzes the architecture and algorithmic characteristics of mainstream ML models. Cross-process comparison reveals that the physical nature of each welding process determines the focus of ML research, model selection, and performance metrics. The study quantitatively compares the task-specific metrics of various models and presents successful industrial application cases. Despite significant progress, challenges persist in constructing high-quality and standardized datasets, improving model interpretability and generalization, and achieving robust real-time control in dynamic industrial environments. Based on the summarized emerging challenges, the perspectives on further development direction of applying ML in intelligent welding are also discussed.

对智能和自主制造日益增长的需求推动了机器学习（ML）与现代焊接工艺的集成。由于它能够模拟非线性和跨尺度的相互作用，并从复杂的高维数据中提取关键特征，机器学习正在迅速改变焊接过程的设计、监测和评估。在此基础上，系统综述了近十年来四种典型焊接工艺（电弧焊、激光焊、电阻焊和搅拌摩擦焊）在机器学习方面的研究进展。首先，典型的焊接任务分为三个领域：焊前设计、焊中监测和焊后质量评估。然后详细阐述了所使用的焊接数据类型及其跨不同任务的输入输出关系，并分析了主流ML模型的架构和算法特征。跨工艺比较表明，每个焊接工艺的物理性质决定了机器学习研究、模型选择和性能指标的重点。该研究定量地比较了各种模型的特定任务度量，并给出了成功的工业应用案例。尽管取得了重大进展，但在构建高质量和标准化数据集、提高模型可解释性和泛化以及在动态工业环境中实现鲁棒实时控制方面仍然存在挑战。在总结当前面临的挑战的基础上，对机器学习在智能焊接领域的进一步发展方向进行了展望。

{"title":"Machine learning applications in welding processes: Progresses and opportunities","authors":"Peihao Geng , Yujun Xia , Zhiqiao Dong , Boxuan Men , Bo Pan , Chenhui Shao , Yongbing Li , Jingjing Li","doi":"10.1016/j.ijmachtools.2025.104344","DOIUrl":"10.1016/j.ijmachtools.2025.104344","url":null,"abstract":"<div><div>The increasing demand for intelligent and autonomous manufacturing has driven the integration of machine learning (ML) into modern welding processes. Owing to its ability to model nonlinear and cross-scale interactions and extract critical features from complex, high-dimensional data, ML is rapidly transforming the design, monitoring, and evaluation of welding processes. Based on this, the paper systematically reviews research progress in ML for four representative welding processes (arc, laser, resistance and friction stir welding) over the past decade. First, typical welding tasks are categorized into three domains: pre-weld design, in-process monitoring, and post-weld quality assessment. It then elaborates on the types of welding data used and their input-output relationships across different tasks and analyzes the architecture and algorithmic characteristics of mainstream ML models. Cross-process comparison reveals that the physical nature of each welding process determines the focus of ML research, model selection, and performance metrics. The study quantitatively compares the task-specific metrics of various models and presents successful industrial application cases. Despite significant progress, challenges persist in constructing high-quality and standardized datasets, improving model interpretability and generalization, and achieving robust real-time control in dynamic industrial environments. Based on the summarized emerging challenges, the perspectives on further development direction of applying ML in intelligent welding are also discussed.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"213 ","pages":"Article 104344"},"PeriodicalIF":18.8,"publicationDate":"2025-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145434318","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Insights into the strengthening mechanisms of titanium alloy treated by electropulsing-assisted ultrasonic nanocrystal surface modification: Process, microstructure, and deformation behavior 电脉冲辅助超声纳米晶表面改性对钛合金强化机理的研究：工艺、微观结构和变形行为

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-10-27 DOI: 10.1016/j.ijmachtools.2025.104343

Yu Zhang , Weidong Zhao , Yixuan Ye , Xiao Jia , Yalin Dong , Han Ding , Jian Wang , Chang Ye

Ultrasonic nanocrystal surface modification (UNSM) properly improves fatigue resistance but shows limited processing efficiency on hard-to-deform alloys such as Ti6Al4V due to their high deformation resistance. This study employs an emerging electropulsing-assisted UNSM (EP-UNSM) approach that integrates pulsed current with conventional UNSM to improve strengthening efficiency and associated surface integrity. However, strain localization-induced nanocrystalline plastic instability may lead to fatigue deterioration under low cycle fatigue regimes. To clarify these aspects, systematic microstructure characterization and mechanical testing were conducted to determine the pathways by which EP-UNSM alters plastic deformation and fatigue strengthening mechanisms. Results reveal that, unlike the planar dislocation slip induced by limited plastic deformation in conventional UNSM, EP-UNSM activates pronounced non-basal dislocations and wavy slip patterns, thereby improving the plasticity and producing a deeper gradient nanostructure layer. Moreover, a dynamic electro-annealing mechanism is proposed that involves the synergistic reconfiguration of metastable dislocations and nanocrystals, forming periodic dislocation cells in EP-UNSM samples, in contrast to the sharp triple-junction grain boundaries observed in conventional UNSM. This microstructure evolution mitigates stress concentration at grain boundaries and enhances compressive residual stress stability, ultimately improving fatigue resistance across all stress regimes. These findings advance understanding of electropulsing-assisted deformation and guide anti-fatigue manufacturing strategies for titanium alloys and other hard-to-deform metals.

超声纳米晶表面改性（UNSM）可有效提高Ti6Al4V等难变形合金的抗疲劳性能，但由于其高抗变形能力，其加工效率有限。本研究采用了一种新兴的电脉冲辅助UNSM （EP-UNSM）方法，该方法将脉冲电流与传统的UNSM相结合，以提高强化效率和相关的表面完整性。然而，应变局部化引起的纳米晶塑性失稳可能导致低周疲劳状态下的疲劳恶化。为了阐明这些方面，进行了系统的微观结构表征和力学测试，以确定EP-UNSM改变塑性变形和疲劳强化机制的途径。结果表明，EP-UNSM与常规UNSM中有限塑性变形引起的平面位错滑移不同，EP-UNSM激活了明显的非基底位错和波状滑移模式，从而提高了塑性并产生了更深的梯度纳米结构层。此外，本文还提出了一种动态电退火机制，该机制涉及亚稳位错和纳米晶体的协同重配置，在EP-UNSM样品中形成周期性位错细胞，而不是传统UNSM中观察到的尖锐的三重结晶界。这种微观结构的演变减轻了晶界的应力集中，增强了压残余应力的稳定性，最终提高了所有应力状态下的抗疲劳能力。这些发现促进了对电脉冲辅助变形的理解，并指导了钛合金和其他难变形金属的抗疲劳制造策略。

{"title":"Insights into the strengthening mechanisms of titanium alloy treated by electropulsing-assisted ultrasonic nanocrystal surface modification: Process, microstructure, and deformation behavior","authors":"Yu Zhang , Weidong Zhao , Yixuan Ye , Xiao Jia , Yalin Dong , Han Ding , Jian Wang , Chang Ye","doi":"10.1016/j.ijmachtools.2025.104343","DOIUrl":"10.1016/j.ijmachtools.2025.104343","url":null,"abstract":"<div><div>Ultrasonic nanocrystal surface modification (UNSM) properly improves fatigue resistance but shows limited processing efficiency on hard-to-deform alloys such as Ti6Al4V due to their high deformation resistance. This study employs an emerging electropulsing-assisted UNSM (EP-UNSM) approach that integrates pulsed current with conventional UNSM to improve strengthening efficiency and associated surface integrity. However, strain localization-induced nanocrystalline plastic instability may lead to fatigue deterioration under low cycle fatigue regimes. To clarify these aspects, systematic microstructure characterization and mechanical testing were conducted to determine the pathways by which EP-UNSM alters plastic deformation and fatigue strengthening mechanisms. Results reveal that, unlike the planar dislocation slip induced by limited plastic deformation in conventional UNSM, EP-UNSM activates pronounced non-basal dislocations and wavy slip patterns, thereby improving the plasticity and producing a deeper gradient nanostructure layer. Moreover, a dynamic electro-annealing mechanism is proposed that involves the synergistic reconfiguration of metastable dislocations and nanocrystals, forming periodic dislocation cells in EP-UNSM samples, in contrast to the sharp triple-junction grain boundaries observed in conventional UNSM. This microstructure evolution mitigates stress concentration at grain boundaries and enhances compressive residual stress stability, ultimately improving fatigue resistance across all stress regimes. These findings advance understanding of electropulsing-assisted deformation and guide anti-fatigue manufacturing strategies for titanium alloys and other hard-to-deform metals.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"213 ","pages":"Article 104343"},"PeriodicalIF":18.8,"publicationDate":"2025-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145396509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Real-time capable feedrate optimization for laser processes with redundant axes via two-stage regularized linear programming 基于两阶段正则线性规划的冗余轴激光加工的实时进给优化

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-10-21 DOI: 10.1016/j.ijmachtools.2025.104342

Haijia Xu , Daniel Kurth , Christoph Hinze , Claudius Horsch , David Hecht , Alexander Verl

Galvanometer scanners offer high dynamics and precision for laser processes, but are limited in their workspace. To expand the workspace, the galvanometer scanner can be integrated into a larger mechanical motion system with redundant axes, including slow mechanical axes and fast scanner axes. While this configuration provides additional degrees of freedom in feedrate planning, conventional Computerized Numerical Control (CNC)-based laser machining systems cannot exploit them effectively, resulting in suboptimal finishing times. This paper introduces the first real-time capable, optimization-based approach to the minimum-time planning problem under motion redundancy, considering limits in redundant axes and toolpath dynamics up to the third order. This is achieved by decoupling the nonlinear problem into two linear problems and introducing a sequential windowing and adaptive scaling strategy, which allows the toolpath to be scaled to arbitrary lengths. Additionally, a new numerical approximation of the transformation between axis and Cartesian coordinates is introduced. This allows for optimization without arc-length parameterization and simplifies the previous toolpath geometry processing. The constraint feasibility and computational efficiency of the proposed optimization method are validated using spline toolpaths. On a desktop PC with single-core execution, the computation time remains well below the actual processing time at around 10 %, showing linear scalability with respect to toolpath length. Experiments on two different laser machines equipped with redundant axes further validate the planning performance and computational robustness when following freeform contours with up to

10 000

constraint checkpoints. Compared to an industrial CNC-guided solution based on S-curve motion profiles, the proposed optimization algorithm reduces the finishing time by around 30 % in experiments with and without jerk constraints.

振镜扫描仪为激光加工提供高动态和精度，但其工作空间有限。为了扩大工作空间，振镜扫描仪可以集成到一个更大的具有冗余轴的机械运动系统中，包括慢速机械轴和快速扫描轴。虽然这种配置为进给速度规划提供了额外的自由度，但传统的基于计算机数控（CNC）的激光加工系统无法有效地利用它们，从而导致次优的精加工时间。本文介绍了第一个实时的，基于优化的方法来解决运动冗余下的最小时间规划问题，考虑了冗余轴和三阶刀具轨迹动力学的限制。这是通过将非线性问题解耦为两个线性问题并引入顺序窗口和自适应缩放策略来实现的，该策略允许将刀具路径缩放到任意长度。此外，还引入了一种新的轴与笛卡尔坐标之间变换的数值近似。这允许在没有弧长参数化的情况下进行优化，并简化了以前的刀具路径几何处理。利用样条刀具轨迹验证了所提优化方法约束的可行性和计算效率。在单核执行的桌面PC上，计算时间保持在10%左右，远低于实际处理时间，显示出与工具路径长度相关的线性可伸缩性。在配备冗余轴的两台不同激光机上的实验进一步验证了该算法在跟踪具有多达10000个约束检查点的自由曲面轮廓时的规划性能和计算鲁棒性。与基于s曲线运动轮廓的工业cnc指导解决方案相比，该优化算法在有和没有抽搐约束的实验中可将加工时间缩短约30%。

{"title":"Real-time capable feedrate optimization for laser processes with redundant axes via two-stage regularized linear programming","authors":"Haijia Xu , Daniel Kurth , Christoph Hinze , Claudius Horsch , David Hecht , Alexander Verl","doi":"10.1016/j.ijmachtools.2025.104342","DOIUrl":"10.1016/j.ijmachtools.2025.104342","url":null,"abstract":"<div><div>Galvanometer scanners offer high dynamics and precision for laser processes, but are limited in their workspace. To expand the workspace, the galvanometer scanner can be integrated into a larger mechanical motion system with redundant axes, including slow mechanical axes and fast scanner axes. While this configuration provides additional degrees of freedom in feedrate planning, conventional Computerized Numerical Control (CNC)-based laser machining systems cannot exploit them effectively, resulting in suboptimal finishing times. This paper introduces the first real-time capable, optimization-based approach to the minimum-time planning problem under motion redundancy, considering limits in redundant axes and toolpath dynamics up to the third order. This is achieved by decoupling the nonlinear problem into two linear problems and introducing a sequential windowing and adaptive scaling strategy, which allows the toolpath to be scaled to arbitrary lengths. Additionally, a new numerical approximation of the transformation between axis and Cartesian coordinates is introduced. This allows for optimization without arc-length parameterization and simplifies the previous toolpath geometry processing. The constraint feasibility and computational efficiency of the proposed optimization method are validated using spline toolpaths. On a desktop PC with single-core execution, the computation time remains well below the actual processing time at around 10 %, showing linear scalability with respect to toolpath length. Experiments on two different laser machines equipped with redundant axes further validate the planning performance and computational robustness when following freeform contours with up to <span><math><mrow><mn>10</mn><mspace></mspace><mn>000</mn></mrow></math></span> constraint checkpoints. Compared to an industrial CNC-guided solution based on S-curve motion profiles, the proposed optimization algorithm reduces the finishing time by around 30 % in experiments with and without jerk constraints.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"213 ","pages":"Article 104342"},"PeriodicalIF":18.8,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145359368","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Quasi-in-situ reconstruction and regulating mechanism of plastic flow in steel friction stir welding 钢搅拌摩擦焊中塑性流动的准原位重构及调节机理

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-10-07 DOI: 10.1016/j.ijmachtools.2025.104341

Yuxuan Li , Mingrun Yu , Shikang Gao , Guangda Sun , Haitao An , Li Zhou

Resolving the mystery of plastic flow in steel friction stir welding (FSW) is critical for process. However, constrained by limitations in flow field techniques and insufficient understanding of the underlying physics, a holistic understanding of plastic flow and its regulating mechanism remains largely empirical. In this study, the material response to the mechanical processing of the FSW tool is reconstructed through a quasi-continuous observation technique. The mechanism of cavity filling, the effective range of tool-workpiece contact states, and the real-time boundary of the shear layer are analyzed. At finer scales, multiple independent vertical components are identified, inducing either unstable periodic flow or mass-balancing effects. These components are characterised as vortex structures. Accordingly, a dynamic model is proposed to specifically elucidate the formation of local vortex structures. The model uses tool–workpiece interaction as the basis for a qualitative description to assess the location of vortex activation, a process that can be semi-quantitatively represented through finite element simulations. The dynamic evolution of the vortex is attributed to the constraining effect of solid-state boundaries on the flow field. The real-time boundary of the shear layer is considered as one form of solid-state boundary, whose constraining effect promotes localised vortex formation. Specifically, the formation of captured vortexes is defined based on the assumption of tool-workpiece interaction and the delineation of shear layer boundaries. Model adaptability is preliminarily verified, and a low-cost method is proposed for capturing previously hidden plastic flows. Across a wide range of process parameters, this model effectively explains plastic flow behaviour. These analyses not only advance a comprehensive knowledge of flow dynamics and associated shear behaviour in steel FSW, but also demonstrate that the proposed dynamic model deepens the fundamental understanding of the complex physical mechanisms during the process. Therefore, this study lays a foundation for optimising welding parameters and supports future academic investigations focused on plastic flow or shear behaviour control.

解决钢搅拌摩擦焊接过程中塑性流动的问题是工艺的关键。然而，由于流场技术的局限性和对基础物理的理解不足，对塑性流动及其调节机制的整体理解在很大程度上仍然是经验主义的。在本研究中，通过准连续观测技术重建了材料对FSW刀具机械加工的响应。分析了型腔充填机理、刀工接触状态有效范围以及剪切层的实时边界。在更细的尺度上，多个独立的垂直分量被识别出来，引起不稳定的周期性流动或质量平衡效应。这些部件的特点是涡结构。据此，提出了一个动力学模型来具体解释局部涡旋结构的形成。该模型使用工具-工件相互作用作为定性描述的基础，以评估涡激活的位置，这一过程可以通过有限元模拟半定量地表示。涡旋的动态演化归因于固体边界对流场的约束作用。剪切层的实时边界被认为是固态边界的一种形式，其约束作用促进了局域涡旋的形成。具体地说，捕获涡的形成是基于刀具-工件相互作用的假设和剪切层边界的描绘来定义的。初步验证了模型的适应性，提出了一种低成本捕获先前隐藏的塑性流的方法。在广泛的工艺参数范围内，该模型有效地解释了塑性流动行为。这些分析不仅促进了对钢搅拌搅拌中流动动力学和相关剪切行为的全面了解，而且还证明了所提出的动态模型加深了对过程中复杂物理机制的基本理解。因此，该研究为优化焊接参数奠定了基础，并为未来专注于塑性流动或剪切行为控制的学术研究提供了支持。

{"title":"Quasi-in-situ reconstruction and regulating mechanism of plastic flow in steel friction stir welding","authors":"Yuxuan Li , Mingrun Yu , Shikang Gao , Guangda Sun , Haitao An , Li Zhou","doi":"10.1016/j.ijmachtools.2025.104341","DOIUrl":"10.1016/j.ijmachtools.2025.104341","url":null,"abstract":"<div><div>Resolving the mystery of plastic flow in steel friction stir welding (FSW) is critical for process. However, constrained by limitations in flow field techniques and insufficient understanding of the underlying physics, a holistic understanding of plastic flow and its regulating mechanism remains largely empirical. In this study, the material response to the mechanical processing of the FSW tool is reconstructed through a quasi-continuous observation technique. The mechanism of cavity filling, the effective range of tool-workpiece contact states, and the real-time boundary of the shear layer are analyzed. At finer scales, multiple independent vertical components are identified, inducing either unstable periodic flow or mass-balancing effects. These components are characterised as vortex structures. Accordingly, a dynamic model is proposed to specifically elucidate the formation of local vortex structures. The model uses tool–workpiece interaction as the basis for a qualitative description to assess the location of vortex activation, a process that can be semi-quantitatively represented through finite element simulations. The dynamic evolution of the vortex is attributed to the constraining effect of solid-state boundaries on the flow field. The real-time boundary of the shear layer is considered as one form of solid-state boundary, whose constraining effect promotes localised vortex formation. Specifically, the formation of captured vortexes is defined based on the assumption of tool-workpiece interaction and the delineation of shear layer boundaries. Model adaptability is preliminarily verified, and a low-cost method is proposed for capturing previously hidden plastic flows. Across a wide range of process parameters, this model effectively explains plastic flow behaviour. These analyses not only advance a comprehensive knowledge of flow dynamics and associated shear behaviour in steel FSW, but also demonstrate that the proposed dynamic model deepens the fundamental understanding of the complex physical mechanisms during the process. Therefore, this study lays a foundation for optimising welding parameters and supports future academic investigations focused on plastic flow or shear behaviour control.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"213 ","pages":"Article 104341"},"PeriodicalIF":18.8,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145247950","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Insights into hydrodynamic self-balancing mechanics in adaptive float machining process for nanometric form error control 自适应浮子加工过程中纳米形状误差控制的流体动力自平衡力学研究

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-09-10 DOI: 10.1016/j.ijmachtools.2025.104333

Fang Han, Jingyuan Wang, Wei Gao, Shuai Wang, Bingchun Jia, Cao-Yang Xue, Weijian Zhang, Bing-Feng Ju, Wule Zhu

With advances in semiconductor and aerospace industries, the demand for components with atomic and close-to-atomic scale accuracy is paramount. As a non-contact method, float machining is showing great potential for excellent surface finishing, where the adaptive fluid film between the tool and workpiece plays an essential role. However, its dynamic self-balancing mechanics remains unexplored, without revealing which, the outcome of extreme accuracy could be hardly touched or controlled. To address this issue, a hydrodynamic coupling dataset driven hydrodynamic model with in-situ force-position sensing approach is proposed. This work presents the first systematic elucidation of the transient processes governing the fluid film-tool interaction as the system achieves a balancing state. Simulation and experiments were conducted to demonstrate the adaptive film evolution route, and its relation to the specific process conditions with precise prediction. Insights into the intermediate states and inherent self-balancing mechanism enable exceptional form control ability, namely deterministic removal of 3 nm in depth across an arbitrary 3 mm region. Subsequently, a sinusoidal hyperbolic freeform with form error within ±2 nm PV over a 5 mm region was fabricated using only a stepper-motor platform, and the underlying process produces an atomically ordered, damage-free subsurface. Furthermore, the capability for non-uniform machining was verified by fabricating compound-eye structures and correcting curved surfaces to nanometric form accuracy.

随着半导体和航空航天工业的进步，对原子和接近原子尺度精度的组件的需求是至关重要的。浮子加工作为一种非接触式加工方法，在实现表面精加工方面显示出巨大的潜力，其中刀具与工件之间的自适应流体膜起着重要的作用。然而，它的动态自平衡机制仍未被探索，没有揭示其中，极端精度的结果几乎无法触摸或控制。为了解决这一问题，提出了一种基于原位力-位置传感方法的水动力耦合数据集驱动的水动力模型。这项工作提出了第一个系统的阐明瞬态过程控制流体膜-工具的相互作用，因为系统达到了平衡状态。通过仿真和实验验证了自适应膜演化路线，并对其与具体工艺条件的关系进行了精确预测。对中间状态和固有的自平衡机制的深入了解使其具有特殊的形状控制能力，即在任意3mm区域内确定去除3nm深度。随后，仅使用步进电机平台就可以在5mm区域内制造出形状误差在±2nm PV内的正弦双曲自由曲面，并且底层工艺产生原子有序，无损伤的亚表面。此外，通过制造复合眼结构和将曲面修正到纳米形状精度，验证了非均匀加工的能力。

{"title":"Insights into hydrodynamic self-balancing mechanics in adaptive float machining process for nanometric form error control","authors":"Fang Han, Jingyuan Wang, Wei Gao, Shuai Wang, Bingchun Jia, Cao-Yang Xue, Weijian Zhang, Bing-Feng Ju, Wule Zhu","doi":"10.1016/j.ijmachtools.2025.104333","DOIUrl":"10.1016/j.ijmachtools.2025.104333","url":null,"abstract":"<div><div>With advances in semiconductor and aerospace industries, the demand for components with atomic and close-to-atomic scale accuracy is paramount. As a non-contact method, float machining is showing great potential for excellent surface finishing, where the adaptive fluid film between the tool and workpiece plays an essential role. However, its dynamic self-balancing mechanics remains unexplored, without revealing which, the outcome of extreme accuracy could be hardly touched or controlled. To address this issue, a hydrodynamic coupling dataset driven hydrodynamic model with in-situ force-position sensing approach is proposed. This work presents the first systematic elucidation of the transient processes governing the fluid film-tool interaction as the system achieves a balancing state. Simulation and experiments were conducted to demonstrate the adaptive film evolution route, and its relation to the specific process conditions with precise prediction. Insights into the intermediate states and inherent self-balancing mechanism enable exceptional form control ability, namely deterministic removal of 3 nm in depth across an arbitrary 3 mm region. Subsequently, a sinusoidal hyperbolic freeform with form error within ±2 nm PV over a 5 mm region was fabricated using only a stepper-motor platform, and the underlying process produces an atomically ordered, damage-free subsurface. Furthermore, the capability for non-uniform machining was verified by fabricating compound-eye structures and correcting curved surfaces to nanometric form accuracy.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"212 ","pages":"Article 104333"},"PeriodicalIF":18.8,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145096895","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A novel in-situ field-assisted powder bed laser fusion using liquid metal enabling microstructure control and strength enhancement of austenitic steel 一种利用液态金属的原位场辅助粉末床激光熔合，可控制奥氏体钢的显微组织并提高其强度

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-09-06 DOI: 10.1016/j.ijmachtools.2025.104334

Xiaoyu Liang , Yurong Wang , Wei Liu , Buwei Xiao , Qingze Liu , Yizhuo Sun , Pengcheng Lv , Huabei Peng , Jun Zhou , Lei Zhang , Feng Lin

The layer-by-layer powder bed additive manufacturing approach, which encapsulates the workpiece in powder during processing, imposes limitations on the integration of in-situ field assistance and enhances production costs. In this work, a novel laser powder bed fusion has been proposed in which the layer-wise accumulated powder bed is replaced by a thin powder layer floating on the liquid Sn. Such a liquid-metal-assisted laser powder bed fusion presents unique advantages: the characteristic thermal history of deposited materials due to high thermal conductivity and fluidity of liquid metals provides greater possibilities for microstructure modulation; the recyclable liquid metal also reduces the need for powder in the forming cylinder and reduces the number of times the powder is reused. Based on the normalized process diagram of liquid-metal-assisted laser powder bed fusion, forming experiments were carried out on the austenitic stainless steels, and the mechanisms underlying the regulation of fine-grain regions were investigated, along with an analysis of the microstructure of this region. Results indicated that the high cooling rate during liquid-metal-assisted laser powder bed fusion led to a finer microstructure and a heterogeneous grain structure ranging from submicron to micron scales in the austenitic stainless steels. The formed heterogeneous austenitic steel exhibits a yield strength surpassing 1.1 GPa and a tensile strength of 1.5 GPa, while retaining an average uniform elongation of 7 %. The in-situ heat treatment principles using liquid metal demonstrated in this work have significant applicability across various additive manufacturing processes and precipitation-hardening alloys.

逐层粉末床增材制造方法在加工过程中将工件封装在粉末中，这限制了现场辅助的集成，并提高了生产成本。本文提出了一种新型的激光粉末床熔合方法，即用悬浮在液态锡上的薄粉末层代替分层堆积的粉末床。这种液体金属辅助激光粉末床熔融具有独特的优势：由于液态金属的高导热性和流动性，沉积材料的特征热历史为微观结构调制提供了更大的可能性；可回收的液态金属还减少了成形气缸对粉末的需求，减少了粉末的重复使用次数。基于液体金属辅助激光粉末床熔合的归一化工艺图，对奥氏体不锈钢进行了成形实验，探讨了细晶区调控的机理，并对该区域的显微组织进行了分析。结果表明，液体金属辅助激光粉末床熔合过程中，高冷却速率使奥氏体不锈钢的显微组织更细，晶粒结构在亚微米到微米之间。形成的非均相奥氏体钢屈服强度超过1.1 GPa，抗拉强度达到1.5 GPa，同时保持7%的平均均匀伸长率。在这项工作中证明的使用液态金属的原位热处理原理在各种增材制造工艺和沉淀硬化合金中具有重要的适用性。

{"title":"A novel in-situ field-assisted powder bed laser fusion using liquid metal enabling microstructure control and strength enhancement of austenitic steel","authors":"Xiaoyu Liang , Yurong Wang , Wei Liu , Buwei Xiao , Qingze Liu , Yizhuo Sun , Pengcheng Lv , Huabei Peng , Jun Zhou , Lei Zhang , Feng Lin","doi":"10.1016/j.ijmachtools.2025.104334","DOIUrl":"10.1016/j.ijmachtools.2025.104334","url":null,"abstract":"<div><div>The layer-by-layer powder bed additive manufacturing approach, which encapsulates the workpiece in powder during processing, imposes limitations on the integration of in-situ field assistance and enhances production costs. In this work, a novel laser powder bed fusion has been proposed in which the layer-wise accumulated powder bed is replaced by a thin powder layer floating on the liquid Sn. Such a liquid-metal-assisted laser powder bed fusion presents unique advantages: the characteristic thermal history of deposited materials due to high thermal conductivity and fluidity of liquid metals provides greater possibilities for microstructure modulation; the recyclable liquid metal also reduces the need for powder in the forming cylinder and reduces the number of times the powder is reused. Based on the normalized process diagram of liquid-metal-assisted laser powder bed fusion, forming experiments were carried out on the austenitic stainless steels, and the mechanisms underlying the regulation of fine-grain regions were investigated, along with an analysis of the microstructure of this region. Results indicated that the high cooling rate during liquid-metal-assisted laser powder bed fusion led to a finer microstructure and a heterogeneous grain structure ranging from submicron to micron scales in the austenitic stainless steels. The formed heterogeneous austenitic steel exhibits a yield strength surpassing 1.1 GPa and a tensile strength of 1.5 GPa, while retaining an average uniform elongation of 7 %. The in-situ heat treatment principles using liquid metal demonstrated in this work have significant applicability across various additive manufacturing processes and precipitation-hardening alloys.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"212 ","pages":"Article 104334"},"PeriodicalIF":18.8,"publicationDate":"2025-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145046792","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

The typical cellular microstructures developed in powder-based additively manufactured metallic materials: formation mechanisms, properties, outlooks and challenges 粉末基增材制造金属材料中典型的细胞微结构：形成机制、性能、前景和挑战

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-09-05 DOI: 10.1016/j.ijmachtools.2025.104332

Jianying Wang , Heng Li , M.W. Fu

Cellular microstructures are intrinsically associated with the printability and mechanical-functional performance of laser-additively manufactured metallic materials. An in-depth understanding of formation mechanisms under extreme processing conditions and impacts on mechanical-functional performance remains critical to improving the application prospects of laser additive manufacturing. In this paper, vital insights into the characteristics, formation mechanisms, mechanical-functional performance, and prospects of cellular microstructures are orchestrated and articulated. First, the differences between dislocation cellular microstructures obtained from conventional methods and those induced by additive manufacturing are summarised through a comparative analysis. Based on the diverse environments of sub-boundaries, almost all cellular microstructures in metallic materials are then exemplified and classified into three categories: dislocation-formed cellular microstructures, both with and without elemental segregation, and eutectic-formed cellular microstructures. For each category, its formation mechanisms related to analysis approaches and evaluation of mechanical-functional performance are delineated and discussed in detail. Finally, insights into the formation mechanisms, model development, thermal stability of cellular microstructures, and countermeasures for aspects of their negative influence on printability and performance are presented. Collectively, this systematic review of cellular microstructures provides a foundational framework to guide the design, manufacture, and industrial-scale implementation of high-performance metallic components.

细胞微结构与激光增材制造的金属材料的可打印性和机械功能性能有着内在的联系。深入了解极端加工条件下的形成机制及其对机械功能性能的影响对于提高激光增材制造的应用前景至关重要。在本文中，对细胞微结构的特征、形成机制、机械功能性能和前景的重要见解进行了编排和阐述。首先，通过对比分析，总结了传统方法获得的位错细胞微观结构与增材制造方法获得的位错细胞微观结构之间的差异。基于不同的亚边界环境，对金属材料中几乎所有的细胞微结构进行了举例，并将其分为三类：位错形成的细胞微结构（包括元素偏析和不存在元素偏析的细胞微结构）和共晶形成的细胞微结构。对于每个类别，其形成机制与分析方法和评估的机械功能性能进行了详细的描述和讨论。最后，介绍了细胞微结构的形成机制、模型开发、热稳定性以及它们对打印性能和性能的负面影响的对策。总的来说，这种对细胞微结构的系统回顾为指导高性能金属部件的设计、制造和工业规模实施提供了基础框架。

{"title":"The typical cellular microstructures developed in powder-based additively manufactured metallic materials: formation mechanisms, properties, outlooks and challenges","authors":"Jianying Wang , Heng Li , M.W. Fu","doi":"10.1016/j.ijmachtools.2025.104332","DOIUrl":"10.1016/j.ijmachtools.2025.104332","url":null,"abstract":"<div><div>Cellular microstructures are intrinsically associated with the printability and mechanical-functional performance of laser-additively manufactured metallic materials. An in-depth understanding of formation mechanisms under extreme processing conditions and impacts on mechanical-functional performance remains critical to improving the application prospects of laser additive manufacturing. In this paper, vital insights into the characteristics, formation mechanisms, mechanical-functional performance, and prospects of cellular microstructures are orchestrated and articulated. First, the differences between dislocation cellular microstructures obtained from conventional methods and those induced by additive manufacturing are summarised through a comparative analysis. Based on the diverse environments of sub-boundaries, almost all cellular microstructures in metallic materials are then exemplified and classified into three categories: dislocation-formed cellular microstructures, both with and without elemental segregation, and eutectic-formed cellular microstructures. For each category, its formation mechanisms related to analysis approaches and evaluation of mechanical-functional performance are delineated and discussed in detail. Finally, insights into the formation mechanisms, model development, thermal stability of cellular microstructures, and countermeasures for aspects of their negative influence on printability and performance are presented. Collectively, this systematic review of cellular microstructures provides a foundational framework to guide the design, manufacture, and industrial-scale implementation of high-performance metallic components.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"212 ","pages":"Article 104332"},"PeriodicalIF":18.8,"publicationDate":"2025-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145046565","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Manufacturing limit and structural scale effect of thin-wall/rod units in lattice structure fabricated by laser powder bed fusion 激光粉末床熔合薄壁/棒晶格结构单元的制造极限及结构规模效应

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-08-28 DOI: 10.1016/j.ijmachtools.2025.104323

Qiao Zhong, Mengxiao Jin, Shihao Bie, Liying Meng, Yisong Wang, Kaiwen Wei, Jiapei Liu, Jianqiang Gong, Yu Yang, Anqi Ouyang, Xiangyou Li, Xiaoyan Zeng

The growing maturity of additive manufacturing (AM) technologies, represented by laser powder bed fusion (LPBF), has greatly facilitated the design and manufacturing of lattice structure. However, the process constraints of the minimum features (thin-wall/rod units) in lattice structure remains poorly understood. This study investigated the manufacturing limits and the relevant failure mechanism of thin-wall/rod units fabricated by LPBF. The effects of structural scale (dimension and inclination angle) on surface morphology, microstructure, and mechanical properties were also studied. Results indicate that the failure of thin-wall/rod units at critical dimension and inclination angle was driven by interlayer molten track mismatch and warping effect in cantilever region, respectively. AM process simulations reveal that rod unit exhibit better manufacturability at small inclination angles compared to thin-wall unit, due to less significant stress deformation. A clear dependence of defect behavior, surface morphology and microstructural characteristics on structural scale is identified. A multi-physics model was created to observe the development of down-skin surface quality of the thin-wall units at small inclination angles. Microstructural analysis reveals transitions between super-refined equiaxed grains, millimeter-long columnar grains, and centimeter-long columnar grains at different scale governed by temperature gradient and grain competitive mechanisms. Additionally, a declining tendency in tensile strength is detected with decreasing feature dimension and inclination angle, primarily because the poor side surface quality and higher porosity accelerated crack initiation and propagation. Moreover, two types of lattice structures were fabricated accordingly. Their characterization results confirm the applicability of the new findings from thin-wall/rod unit experiments, providing insights for the efficient manufacturing of lattice structures endowed with excellent performance.

以激光粉末床熔融（LPBF）为代表的增材制造（AM）技术日益成熟，极大地促进了晶格结构的设计和制造。然而，晶格结构中最小特征（薄壁/棒状单元）的工艺约束仍然知之甚少。研究了LPBF薄壁杆件的制造极限及其失效机理。研究了结构尺度（尺寸和倾角）对表面形貌、显微组织和力学性能的影响。结果表明，薄壁/杆单元在临界尺寸和倾斜角度下的失效分别由层间熔迹失配和悬臂区域翘曲效应驱动。增材制造过程模拟表明，与薄壁单元相比，棒材单元在小倾角下具有更好的可制造性，因为应力变形较小。缺陷行为、表面形貌和微观结构特征对结构尺度有明显的依赖性。建立了多物理场模型，观察了小倾角下薄壁单元的下表皮质量的变化。显微组织分析揭示了不同尺度下超细化等轴晶、毫米长柱状晶和厘米长柱状晶之间的转变受温度梯度和晶粒竞争机制的控制。拉伸强度随特征维数和倾角的减小呈下降趋势，主要原因是侧表面质量差、孔隙率高加速了裂纹的萌生和扩展。并据此制作了两种类型的晶格结构。他们的表征结果证实了薄壁/棒单元实验新发现的适用性，为高效制造具有优异性能的晶格结构提供了见解。

{"title":"Manufacturing limit and structural scale effect of thin-wall/rod units in lattice structure fabricated by laser powder bed fusion","authors":"Qiao Zhong, Mengxiao Jin, Shihao Bie, Liying Meng, Yisong Wang, Kaiwen Wei, Jiapei Liu, Jianqiang Gong, Yu Yang, Anqi Ouyang, Xiangyou Li, Xiaoyan Zeng","doi":"10.1016/j.ijmachtools.2025.104323","DOIUrl":"10.1016/j.ijmachtools.2025.104323","url":null,"abstract":"<div><div>The growing maturity of additive manufacturing (AM) technologies, represented by laser powder bed fusion (LPBF), has greatly facilitated the design and manufacturing of lattice structure. However, the process constraints of the minimum features (thin-wall/rod units) in lattice structure remains poorly understood. This study investigated the manufacturing limits and the relevant failure mechanism of thin-wall/rod units fabricated by LPBF. The effects of structural scale (dimension and inclination angle) on surface morphology, microstructure, and mechanical properties were also studied. Results indicate that the failure of thin-wall/rod units at critical dimension and inclination angle was driven by interlayer molten track mismatch and warping effect in cantilever region, respectively. AM process simulations reveal that rod unit exhibit better manufacturability at small inclination angles compared to thin-wall unit, due to less significant stress deformation. A clear dependence of defect behavior, surface morphology and microstructural characteristics on structural scale is identified. A multi-physics model was created to observe the development of down-skin surface quality of the thin-wall units at small inclination angles. Microstructural analysis reveals transitions between super-refined equiaxed grains, millimeter-long columnar grains, and centimeter-long columnar grains at different scale governed by temperature gradient and grain competitive mechanisms. Additionally, a declining tendency in tensile strength is detected with decreasing feature dimension and inclination angle, primarily because the poor side surface quality and higher porosity accelerated crack initiation and propagation. Moreover, two types of lattice structures were fabricated accordingly. Their characterization results confirm the applicability of the new findings from thin-wall/rod unit experiments, providing insights for the efficient manufacturing of lattice structures endowed with excellent performance.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"212 ","pages":"Article 104323"},"PeriodicalIF":18.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144989499","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Derivation of physical equations for high-speed laser welding using large language models 用大语言模型推导高速激光焊接物理方程

IF 18.8 1区工程技术 Q1 ENGINEERING, MANUFACTURING

International Journal of Machine Tools & Manufacture

Pub Date : 2025-08-20 DOI: 10.1016/j.ijmachtools.2025.104320

Kyubok Lee , Zhengxiao Yu , Zen-Hao Lai , Peihao Geng , Teresa J. Rinker , Changbai Tan , Blair Carlson , Siguang Xu , Jingjing Li

It is challenging to formulate complex physical phenomena that occur in a manufacturing process, particularly when the available data are limited, rendering conventional data-driven approaches ineffective. This study aims to predict humping onset in high-speed laser welding by introducing a novel framework, namely text-to-equations generative pre-trained transformer (T2EGPT). This method leverages the capabilities of large language models (LLMs), in combination with sparse experimental data and enriched literature data, to derive an interpretable and generalizable equation for predicting humping initiation. By capturing key correlations among physical parameters, T2EGPT generates a compact and dimensionless expression that accurately predicts hump formation. The equation reveals that humping arises from the interplay between inertia-driven backward melt flow and capillary-driven surface stabilization, where inertial forces drive molten metal backward and capillary forces resist surface deformation. Compared to traditional data-driven models, T2EGPT demonstrates enhanced predictive accuracy and cross-material transferability. More broadly, this study highlights the potential of LLMs to integrate textual information with data-driven discovery, enabling the extraction of physical laws in data-scarce scientific domains.

对于制造过程中发生的复杂物理现象，特别是在可用数据有限的情况下，传统的数据驱动方法是无效的。本研究旨在通过引入一个新的框架，即文本-方程生成预训练变压器（T2EGPT）来预测高速激光焊接中的驼峰发生。该方法利用大型语言模型（llm）的能力，结合稀疏的实验数据和丰富的文献数据，推导出一个可解释和可推广的预测驼峰起始的方程。通过捕获物理参数之间的关键相关性，T2EGPT生成一个紧凑的无量纲表达式，可以准确预测驼峰的形成。由方程可知，驼峰是由惯性力驱动的熔体向后流动和毛细力驱动的表面稳定相互作用产生的，其中惯性力驱动金属液向后流动，毛细力抑制表面变形。与传统的数据驱动模型相比，T2EGPT具有更高的预测准确性和跨材料可移植性。更广泛地说，这项研究强调了法学硕士将文本信息与数据驱动的发现相结合的潜力，从而能够在数据稀缺的科学领域中提取物理定律。

{"title":"Derivation of physical equations for high-speed laser welding using large language models","authors":"Kyubok Lee , Zhengxiao Yu , Zen-Hao Lai , Peihao Geng , Teresa J. Rinker , Changbai Tan , Blair Carlson , Siguang Xu , Jingjing Li","doi":"10.1016/j.ijmachtools.2025.104320","DOIUrl":"10.1016/j.ijmachtools.2025.104320","url":null,"abstract":"<div><div>It is challenging to formulate complex physical phenomena that occur in a manufacturing process, particularly when the available data are limited, rendering conventional data-driven approaches ineffective. This study aims to predict humping onset in high-speed laser welding by introducing a novel framework, namely text-to-equations generative pre-trained transformer (T2EGPT). This method leverages the capabilities of large language models (LLMs), in combination with sparse experimental data and enriched literature data, to derive an interpretable and generalizable equation for predicting humping initiation. By capturing key correlations among physical parameters, T2EGPT generates a compact and dimensionless expression that accurately predicts hump formation. The equation reveals that humping arises from the interplay between inertia-driven backward melt flow and capillary-driven surface stabilization, where inertial forces drive molten metal backward and capillary forces resist surface deformation. Compared to traditional data-driven models, T2EGPT demonstrates enhanced predictive accuracy and cross-material transferability. More broadly, this study highlights the potential of LLMs to integrate textual information with data-driven discovery, enabling the extraction of physical laws in data-scarce scientific domains.</div></div>","PeriodicalId":14011,"journal":{"name":"International Journal of Machine Tools & Manufacture","volume":"211 ","pages":"Article 104320"},"PeriodicalIF":18.8,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144892072","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0