Foreign object damage (FOD) induced fatigue strength attenuation of aeroengine blades is a critical challenge in reliability design, as it is affected by the coupling of multiple factors such as foreign object characteristics, impact angle, and damage morphology. Experimental data in this context often suffer from limited sample size, significant noise, and incomplete feature coverage. Traditional physical empirical methods rely on simplified assumptions, struggling to adapt to nonlinear evolution of complex damage, while pure data-driven models lack physical constraints, tending to overfit or produce physically implausible predictions. To address these limitations, this study proposes a physics-informed enhanced machine learning method for FOD fatigue strength prediction. The core of physical information enhanced-XGBoost (PIE-XGBoost) lies in embedding physical priors from the Peterson empirical formula into the XGBoost loss function and precomputing theoretical fatigue strength using parameters such as damage depth and foreign object diameter as physical constraints throughout training. Additionally, an adaptive physical constraint strength mechanism is introduced to dynamically adjust regularization coefficients via training error, balancing physical constraint guidance in the early stages with data-driven optimization in the later stages. Finally, based on the simulated blade experimental data verification and analysis of FOD blades, the average error of PIE-XGBoost is 3.2%. Compared with the traditional physical empirical formula’s average error of 41.47%, PIE-XGBoost reduces the error by 38.27%, thus verifying the effectiveness of the method. Further application of this method to actual aeroengine blades can provide technical support for aeroengine maintenance, which has high engineering practical significance and application prospects.
{"title":"Physical information-enhanced machine learning method for high cycle fatigue strength prediction of foreign object damaged aeroengine blades","authors":"Yuming Huang, Yibo Shang, Yu Fu, Chen Wang, Qiang Chen, Yun He, Qingyang Shen, Weisi Gao, Shifeng Wen, Weifeng He, Ming Li, Zhifen Zhang, Liucheng Zhou, Zhenhua Zhao","doi":"10.1016/j.ijfatigue.2026.109540","DOIUrl":"https://doi.org/10.1016/j.ijfatigue.2026.109540","url":null,"abstract":"Foreign object damage (FOD) induced fatigue strength attenuation of aeroengine blades is a critical challenge in reliability design, as it is affected by the coupling of multiple factors such as foreign object characteristics, impact angle, and damage morphology. Experimental data in this context often suffer from limited sample size, significant noise, and incomplete feature coverage. Traditional physical empirical methods rely on simplified assumptions, struggling to adapt to nonlinear evolution of complex damage, while pure data-driven models lack physical constraints, tending to overfit or produce physically implausible predictions. To address these limitations, this study proposes a physics-informed enhanced machine learning method for FOD fatigue strength prediction. The core of physical information enhanced-XGBoost (PIE-XGBoost) lies in embedding physical priors from the Peterson empirical formula into the XGBoost loss function and precomputing theoretical fatigue strength using parameters such as damage depth and foreign object diameter as physical constraints throughout training. Additionally, an adaptive physical constraint strength mechanism is introduced to dynamically adjust regularization coefficients via training error, balancing physical constraint guidance in the early stages with data-driven optimization in the later stages. Finally, based on the simulated blade experimental data verification and analysis of FOD blades, the average error of PIE-XGBoost is 3.2%. Compared with the traditional physical empirical formula’s average error of 41.47%, PIE-XGBoost reduces the error by 38.27%, thus verifying the effectiveness of the method. Further application of this method to actual aeroengine blades can provide technical support for aeroengine maintenance, which has high engineering practical significance and application prospects.","PeriodicalId":14112,"journal":{"name":"International Journal of Fatigue","volume":"5 1","pages":""},"PeriodicalIF":6.0,"publicationDate":"2026-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146146769","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-09DOI: 10.1016/j.ijfatigue.2026.109559
Niklas Sayer-Duffhauß, Markus Fried, Sebastian Münstermann
With increasing capabilities in non-destructive testing methods, image-based modeling of porosity becomes more widely adopted. While some empirically- and fracture-mechanics-derived criteria exist in literature and standards, image-based models would benefit from a more thorough assessment of porosity interaction. In this paper we model the stress response of pairs of pores in a vast number of possible spatial configurations, using the example of the alloy MAR-M247 and real casting pores, whose morphologies were acquired using computed tomography (CT). Lastly, we compute the interaction distance and derive criteria for non-interaction for two popular types of image-based fatigue models (non-local and global), based on the size of the potentially interacting pores.
{"title":"Studying porosity interaction of real casting pore morphologies in image-based models during fatigue loading","authors":"Niklas Sayer-Duffhauß, Markus Fried, Sebastian Münstermann","doi":"10.1016/j.ijfatigue.2026.109559","DOIUrl":"https://doi.org/10.1016/j.ijfatigue.2026.109559","url":null,"abstract":"With increasing capabilities in non-destructive testing methods, image-based modeling of porosity becomes more widely adopted. While some empirically- and fracture-mechanics-derived criteria exist in literature and standards, image-based models would benefit from a more thorough assessment of porosity interaction. In this paper we model the stress response of pairs of pores in a vast number of possible spatial configurations, using the example of the alloy MAR-M247 and real casting pores, whose morphologies were acquired using computed tomography (CT). Lastly, we compute the interaction distance and derive criteria for non-interaction for two popular types of image-based fatigue models (non-local and global), based on the size of the potentially interacting pores.","PeriodicalId":14112,"journal":{"name":"International Journal of Fatigue","volume":"35 1","pages":""},"PeriodicalIF":6.0,"publicationDate":"2026-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146146768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-08DOI: 10.1016/j.ijfatigue.2026.109557
Yang Xin-Yi, Zhu Li-Na, Xu Zhong-Wei, Wang Xi-Shu
{"title":"Probabilistic evaluation on fatigue small cracking characteristics of light metallic alloys under in-situ SEM fatigue tests using the weakest link theory","authors":"Yang Xin-Yi, Zhu Li-Na, Xu Zhong-Wei, Wang Xi-Shu","doi":"10.1016/j.ijfatigue.2026.109557","DOIUrl":"https://doi.org/10.1016/j.ijfatigue.2026.109557","url":null,"abstract":"","PeriodicalId":14112,"journal":{"name":"International Journal of Fatigue","volume":"72 1","pages":""},"PeriodicalIF":6.0,"publicationDate":"2026-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146138671","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-07DOI: 10.1016/j.ijfatigue.2026.109553
Han Yan, Dawei Huang, Aofei Li, Zhenyu He, Heming Xu, Naixian Hou, Xiaojun Yan
{"title":"Experimental investigation and modeling of the superalloy crack growth behavior under combined high and low cycle fatigue","authors":"Han Yan, Dawei Huang, Aofei Li, Zhenyu He, Heming Xu, Naixian Hou, Xiaojun Yan","doi":"10.1016/j.ijfatigue.2026.109553","DOIUrl":"https://doi.org/10.1016/j.ijfatigue.2026.109553","url":null,"abstract":"","PeriodicalId":14112,"journal":{"name":"International Journal of Fatigue","volume":"17 1","pages":""},"PeriodicalIF":6.0,"publicationDate":"2026-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146135261","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-06DOI: 10.1016/j.ijfatigue.2026.109552
Ivo Šulák, Markéta Gálíková, Tomáš Babinský, Ladislav Poczklán, Ivo Kuběna, Stefan Guth
{"title":"Thermomechanical fatigue performance of additively manufactured Inconel 939","authors":"Ivo Šulák, Markéta Gálíková, Tomáš Babinský, Ladislav Poczklán, Ivo Kuběna, Stefan Guth","doi":"10.1016/j.ijfatigue.2026.109552","DOIUrl":"https://doi.org/10.1016/j.ijfatigue.2026.109552","url":null,"abstract":"","PeriodicalId":14112,"journal":{"name":"International Journal of Fatigue","volume":"89 1","pages":""},"PeriodicalIF":6.0,"publicationDate":"2026-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146135262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-05DOI: 10.1016/j.ijfatigue.2026.109535
Samira Ghadar, Ali Fatemi
{"title":"A review of fretting fatigue life prediction models and application of the critical plane approach to selected literature datasets","authors":"Samira Ghadar, Ali Fatemi","doi":"10.1016/j.ijfatigue.2026.109535","DOIUrl":"https://doi.org/10.1016/j.ijfatigue.2026.109535","url":null,"abstract":"","PeriodicalId":14112,"journal":{"name":"International Journal of Fatigue","volume":"47 1","pages":""},"PeriodicalIF":6.0,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146134457","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：