Machine Learning for Fatigue Life Prediction of Additively Manufactured Metals: A Review
Shuyan Wang, Yitong Lv, Libo Fan, Yu Liu
Xinxiang Institute of Engineering University of Wollongong
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
This review examines machine-learning (ML) approaches for fatigue life prediction in additively manufactured (AM) metals, where fatigue is governed by the coupled effects of defects, surface condition, residual stress, and microstructure. To organize the existing literature, the input feature space is grouped into four categories: process features, material and microstructural features, defect and surface-state features, and loading and environmental features. Recent studies are reviewed with emphasis on modeling frameworks, feature organization, validation strategies, and the roles of defects and post-processing. Particular attention is given to the cross-dataset comparability of reported model performance and to the rigor of current evaluation protocols. Current progress is constrained by inconsistent data reporting, small sample sizes, incomplete metadata, insufficient grouped validation, limited comparability across studies, and poor interpretability. Future work should prioritize standardized benchmark datasets, physics-informed and probabilistic models, defect-sensitive feature representation, and reliability-oriented evaluation supported by improved metadata.
逐年被引趋势
暂无年度引用数据
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
工程Additive Manufacturing Materials and Processes
Fatigue and fracture mechanics · Machine Learning in Materials Science