Machine Learning-Based Strength Prediction for Two-Stage Aged 7050 Aluminum Alloy Forgings in Aircraft Main Support Joints

IF 3.4 3区 材料科学 Q2 MATERIALS SCIENCE, MULTIDISCIPLINARY
Yongjie Liu, Yuanzhi Qian, Weijiu Huang, Xiaofei Zhu, Xusheng Yang, Lingfei Cao, Yanzheng Guo, Mofan Liu, Wenya Xiao, Ke Gan
{"title":"Machine Learning-Based Strength Prediction for Two-Stage Aged 7050 Aluminum Alloy Forgings in Aircraft Main Support Joints","authors":"Yongjie Liu,&nbsp;Yuanzhi Qian,&nbsp;Weijiu Huang,&nbsp;Xiaofei Zhu,&nbsp;Xusheng Yang,&nbsp;Lingfei Cao,&nbsp;Yanzheng Guo,&nbsp;Mofan Liu,&nbsp;Wenya Xiao,&nbsp;Ke Gan","doi":"10.1002/adem.202402024","DOIUrl":null,"url":null,"abstract":"<p>Aluminum alloys, widely regarded as lightweight structural materials, are extensively used in the aerospace industry. The aging process is essential for reducing residual stresses and ensuring alloys quality. Traditional methods for optimizing aging are often time-consuming and expensive. In contrast, machine learning (ML) accelerates material design and performance prediction, significantly minimizing the need for extensive experimentation. In this study, the 7050 aluminum alloy forgings in aircraft main support joints are selected as the research object. A forward prediction model is developed using common ML algorithms, incorporating two-stage aging process parameters and microstructural features as inputs, with yield strength (YS) and ultimate tensile strength (UTS) as outputs. The results demonstrate that the extreme gradient boosting regression model is the most effective for predicting the strength of aluminum alloys, with <i>R</i><sup>2</sup> values exceeding 0.7. By the Shapley additive explanation (SHAP) method and microscopic morphology analysis, the second-stage aging time (<i>t</i><sub>2</sub>) significantly influences YS and UTS. Hence, <i>t</i><sub>2</sub> was selected as the output for constructing the reverse classification model. The support vector machine classification model exhibits optimal performance, attaining macro-accuracy and macro-recall rates of 0.91 and 0.90, respectively.</p>","PeriodicalId":7275,"journal":{"name":"Advanced Engineering Materials","volume":"26 24","pages":""},"PeriodicalIF":3.4000,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Advanced Engineering Materials","FirstCategoryId":"88","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/adem.202402024","RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"MATERIALS SCIENCE, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0

Abstract

Aluminum alloys, widely regarded as lightweight structural materials, are extensively used in the aerospace industry. The aging process is essential for reducing residual stresses and ensuring alloys quality. Traditional methods for optimizing aging are often time-consuming and expensive. In contrast, machine learning (ML) accelerates material design and performance prediction, significantly minimizing the need for extensive experimentation. In this study, the 7050 aluminum alloy forgings in aircraft main support joints are selected as the research object. A forward prediction model is developed using common ML algorithms, incorporating two-stage aging process parameters and microstructural features as inputs, with yield strength (YS) and ultimate tensile strength (UTS) as outputs. The results demonstrate that the extreme gradient boosting regression model is the most effective for predicting the strength of aluminum alloys, with R2 values exceeding 0.7. By the Shapley additive explanation (SHAP) method and microscopic morphology analysis, the second-stage aging time (t2) significantly influences YS and UTS. Hence, t2 was selected as the output for constructing the reverse classification model. The support vector machine classification model exhibits optimal performance, attaining macro-accuracy and macro-recall rates of 0.91 and 0.90, respectively.

基于机器学习的飞机主支承接头两段时效7050铝合金锻件强度预测
铝合金被广泛认为是轻质结构材料,在航空航天工业中得到了广泛的应用。时效过程是降低残余应力、保证合金质量的重要手段。优化老化的传统方法通常既耗时又昂贵。相比之下,机器学习(ML)加速了材料设计和性能预测,大大减少了对大量实验的需求。本研究以7050铝合金飞机主支承接头锻件为研究对象。采用通用ML算法建立了前向预测模型,将两阶段时效过程参数和微观结构特征作为输入,以屈服强度(YS)和极限拉伸强度(UTS)作为输出。结果表明,极值梯度增强回归模型预测铝合金强度最有效,R2值均大于0.7。通过Shapley加性解释(SHAP)方法和微观形貌分析,第二阶段时效时间(t2)对YS和UTS有显著影响。因此,选择t2作为构建反向分类模型的输出。支持向量机分类模型表现出最优的性能,宏正确率和宏召回率分别达到0.91和0.90。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Advanced Engineering Materials
Advanced Engineering Materials 工程技术-材料科学:综合
CiteScore
5.70
自引率
5.60%
发文量
544
审稿时长
1.7 months
期刊介绍: Advanced Engineering Materials is the membership journal of three leading European Materials Societies - German Materials Society/DGM, - French Materials Society/SF2M, - Swiss Materials Federation/SVMT.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信