Pub Date : 2026-04-01Epub Date: 2026-01-16DOI: 10.1016/j.engfailanal.2026.110587
Hyeok-Jun Kwon , Hongseok Kim , Youngchan Kim , Dooyoul Lee
This study analyzed two extremely rare internal object damage (IOD) events in the prevalent J85 engine, which are attributable to the long-term operation of its compressor parts. In both cases, the damaged parts (retaining ring, compressor 7th stator vane) were made of the same alloy 718 material, and the presence of material defects was investigated first. The first case involved a retaining ring fasten pin that fractured in half, collided, and stuck to the rotor blade’s leading edge. The root cause of pin failure was confirmed as a microstructural defect. The second case, a compressor 7th stator vane fracture, was linked to resonance that depleted its fatigue life and led to premature failure. In particular, changes in vibration modes occurring during aeroengine operation, along with microstructural defects, might be the main causes of fatigue crack initiation and propagation. Each unique IOD case can provide valuable insights into advanced gas turbine aeroengine design and operation.
{"title":"Failure analyses of J85 engine compressor caused by Alloy 718 internal object debris","authors":"Hyeok-Jun Kwon , Hongseok Kim , Youngchan Kim , Dooyoul Lee","doi":"10.1016/j.engfailanal.2026.110587","DOIUrl":"10.1016/j.engfailanal.2026.110587","url":null,"abstract":"<div><div>This study analyzed two extremely rare internal object damage (IOD) events in the prevalent J85 engine, which are attributable to the long-term operation of its compressor parts. In both cases, the damaged parts (retaining ring, compressor 7th stator vane) were made of the same alloy 718 material, and the presence of material defects was investigated first. The first case involved a retaining ring fasten pin that fractured in half, collided, and stuck to the rotor blade’s leading edge. The root cause of pin failure was confirmed as a microstructural defect. The second case, a compressor 7th stator vane fracture, was linked to resonance that depleted its fatigue life and led to premature failure. In particular, changes in vibration modes occurring during aeroengine operation, along with microstructural defects, might be the main causes of fatigue crack initiation and propagation. Each unique IOD case can provide valuable insights into advanced gas turbine aeroengine design and operation.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110587"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146001772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2026-01-23DOI: 10.1016/j.engfailanal.2026.110608
V.M. Sreedevi , A. Anisha , Robin Davis , Sujith Mangalathu , Prateek Negi
Accurate prediction of failure is essential for maintaining structural integrity and achieving design efficiency, as it helps prevent catastrophic failures. With the increasing adoption of cold-formed steel (CFS) members in construction, precise estimation of their failure load is necessary, especially as it undergoes various failure modes like local, distortional, global buckling or a combination of these. Existing design standards originally developed for conventional CFS members are not intended for the high strength cold formed steel (HSCFS) members. Present study proposes a hybrid data driven methodology to develop a Machine Learning based High-Fidelity Model (MLHFM) for failure load prediction. The proposed approach is found to be performing well for the failure prediction of high strength cold formed steel square hollow section (HSCFS-SHS) columns. In this hybrid method, twelve experimental data regarding HSCFS-SHS columns are collected, numerical models are generated for the same and machine learning models are developed using data generated from the numerical models. Twelve machine learning (ML) techniques with their tuned hyper-parameters are utilized in present study for developing MLHFM. CatBoost is identified as the best performing MLHFM with the R2, RMSE, MAE and MAPE values of 0.974, 0.033, 0.008 and 0.024 respectively. Additionally, a SHAP (SHapley Additive exPlanations) analysis is performed to interpret the model’s predictions. The adequacy of the developed MLHFM is established by comparing their predictions with experimental results and international design codes. Further, a reliability analysis conducted as per AISI S100 shows that MLHFM prediction is able to achieve a target reliability index of 2.5 (2.85 and 2.61 for resistance factors of 0.8 and 0.85 respectively). Finally, a graphical user interface is established for the failure prediction of HSCFS-SHS column.
{"title":"Innovative hybrid data-driven approach for failure prediction of cold-formed steel columns using high-fidelity models – performance comparison with international design codes","authors":"V.M. Sreedevi , A. Anisha , Robin Davis , Sujith Mangalathu , Prateek Negi","doi":"10.1016/j.engfailanal.2026.110608","DOIUrl":"10.1016/j.engfailanal.2026.110608","url":null,"abstract":"<div><div>Accurate prediction of failure is essential for maintaining structural integrity and achieving design efficiency, as it helps prevent catastrophic failures. With the increasing adoption of cold-formed steel (CFS) members in construction, precise estimation of their failure load is necessary, especially as it undergoes various failure modes like local, distortional, global buckling or a combination of these. Existing design standards originally developed for conventional CFS members are not intended for the high strength cold formed steel (HSCFS) members. Present study proposes a hybrid data driven methodology to develop a Machine Learning based High-Fidelity Model (MLHFM) for failure load prediction. The proposed approach is found to be performing well for the failure prediction of high strength cold formed steel square hollow section (HSCFS-SHS) columns. In this hybrid method, twelve experimental data regarding HSCFS-SHS columns are collected, numerical models are generated for the same and machine learning models are developed using data generated from the numerical models. Twelve machine learning (ML) techniques with their tuned hyper-parameters are utilized in present study for developing MLHFM. CatBoost is identified as the best performing MLHFM with the <em>R</em><sup>2</sup>, RMSE, MAE and MAPE values of 0.974, 0.033, 0.008 and 0.024 respectively. Additionally, a SHAP (SHapley Additive exPlanations) analysis is performed to interpret the model’s predictions. The adequacy of the developed MLHFM is established by comparing their predictions with experimental results and international design codes. Further, a reliability analysis conducted as per AISI S100 shows that MLHFM prediction is able to achieve a target reliability index of 2.5 (2.85 and 2.61 for resistance factors of 0.8 and 0.85 respectively). Finally, a graphical user interface is established for the failure prediction of HSCFS-SHS column.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110608"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075744","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2026-01-27DOI: 10.1016/j.engfailanal.2026.110624
Zhibin Zhao , Nian Wang , Jianwu Zhou , Haijun Jiang , Heyang Miao , Wei Zhang , Zhengwei Yang
With the expanding application of thermoplastic composites in aerospace engineering, there is an urgent need for non-destructive testing of impact damage in CF/PEEK composites. This study aims to address two major challenges in ultrasonic infrared thermography (UIT) testing of CF/PEEK composites: unclear excitation parameter settings and low damage segmentation accuracy. To this end, a two-stage method combining excitation parameter optimization and unsupervised thermal image segmentation is proposed. First, a quantitative mapping model between excitation parameters and thermal response characteristics was systematically established through single-factor experiments and response surface methodology, thereby determining the optimal excitation parameter combination. On this basis, a novel multi-order graph clustering segmentation model named Fast Multi-order Graph Fusion Clustering (FMGFC) was developed. Its core innovations lie in the introduction of an adaptive multi-order graph selection mechanism and a graph fusion strategy based on low-rank tensor approximation. The model generates anchors through superpixel learning, reducing computational complexity to linear level; it then adaptively selects specific orders of sample-anchor graphs and learns their consistent and accurate similarity representations via low-rank tensor approximation. Experimental results show that the optimized excitation scheme significantly enhances the signal-to-noise ratio of damage thermal images; the proposed FMGFC model achieved over 95% segmentation accuracy on three specimens while reducing runtime by approximately 50–70% compared to other algorithms, effectively unifying high precision and high efficiency. This research provides critical technical support and methodological references for UIT testing of thermoplastic composites.
{"title":"CF/PEEK composites ultrasonic infrared thermography testing: Excitation parameter optimization and fast multi-order graph fusion clustering segmentation","authors":"Zhibin Zhao , Nian Wang , Jianwu Zhou , Haijun Jiang , Heyang Miao , Wei Zhang , Zhengwei Yang","doi":"10.1016/j.engfailanal.2026.110624","DOIUrl":"10.1016/j.engfailanal.2026.110624","url":null,"abstract":"<div><div>With the expanding application of thermoplastic composites in aerospace engineering, there is an urgent need for non-destructive testing of impact damage in CF/PEEK composites. This study aims to address two major challenges in ultrasonic infrared thermography (UIT) testing of CF/PEEK composites: unclear excitation parameter settings and low damage segmentation accuracy. To this end, a two-stage method combining excitation parameter optimization and unsupervised thermal image segmentation is proposed. First, a quantitative mapping model between excitation parameters and thermal response characteristics was systematically established through single-factor experiments and response surface methodology, thereby determining the optimal excitation parameter combination. On this basis, a novel multi-order graph clustering segmentation model named Fast Multi-order Graph Fusion Clustering (FMGFC) was developed. Its core innovations lie in the introduction of an adaptive multi-order graph selection mechanism and a graph fusion strategy based on low-rank tensor approximation. The model generates anchors through superpixel learning, reducing computational complexity to linear level; it then adaptively selects specific orders of sample-anchor graphs and learns their consistent and accurate similarity representations via low-rank tensor approximation. Experimental results show that the optimized excitation scheme significantly enhances the signal-to-noise ratio of damage thermal images; the proposed FMGFC model achieved over 95% segmentation accuracy on three specimens while reducing runtime by approximately 50–70% compared to other algorithms, effectively unifying high precision and high efficiency. This research provides critical technical support and methodological references for UIT testing of thermoplastic composites.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110624"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075751","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2026-01-23DOI: 10.1016/j.engfailanal.2026.110606
Wei Lan , Chang Cong , Qingjun Gong , Bin Wang , Wuxi Bi , Daoqing Liu , Chengwei Xu , Zhe Wang
The accompanying optical cable, a critical conduit for communication and data transmission in oil and gas pipelines, plays a vital role in pipeline integrity management. However, extreme weather conditions, particularly lightning strikes, pose significant threats to the safe operation of both pipelines and their accompanying optical cables. In this work, the failure of accompanying optical cables caused by a lightning strike in Inner Mongolia, China, on April 13, 2024, is analyzed through laboratory tests on the lightning breakdown of accompanying optical cables and numerical simulations of pipeline lightning strikes, and specific protective strategies are proposed. According to the tested and simulated results, the direct cause of the event is identified as the location of the accompanying optical cable being within the soil ionization radius. The leading cause is high soil resistivity, and the impulse voltage of the lightning on the accompanying optical cable exceeding its breakdown voltage threshold. Based on the causes and characteristics of the actual lightning strike failure accident involving the accompanying optical cables, protective measures are proposed, prioritizing inner inspection of the pipelines, investigation of the trees near the pipelines, and the pipelines in high lightning strike areas. This work provides essential methods and preventive measures for ensuring pipeline integrity management and safe operation throughout the practical design and production management process of pipelines.
{"title":"Lightning-induced failure mechanisms of co-located pipeline optical cables: a soil ionization modeling approach","authors":"Wei Lan , Chang Cong , Qingjun Gong , Bin Wang , Wuxi Bi , Daoqing Liu , Chengwei Xu , Zhe Wang","doi":"10.1016/j.engfailanal.2026.110606","DOIUrl":"10.1016/j.engfailanal.2026.110606","url":null,"abstract":"<div><div>The accompanying optical cable, a critical conduit for communication and data transmission in oil and gas pipelines, plays a vital role in pipeline integrity management. However, extreme weather conditions, particularly lightning strikes, pose significant threats to the safe operation of both pipelines and their accompanying optical cables. In this work, the failure of accompanying optical cables caused by a lightning strike in Inner Mongolia, China, on April 13, 2024, is analyzed through laboratory tests on the lightning breakdown of accompanying optical cables and numerical simulations of pipeline lightning strikes, and specific protective strategies are proposed. According to the tested and simulated results, the direct cause of the event is identified as the location of the accompanying optical cable being within the soil ionization radius. The leading cause is high soil resistivity, and the impulse voltage of the lightning on the accompanying optical cable exceeding its breakdown voltage threshold. Based on the causes and characteristics of the actual lightning strike failure accident involving the accompanying optical cables, protective measures are proposed, prioritizing inner inspection of the pipelines, investigation of the trees near the pipelines, and the pipelines in high lightning strike areas. This work provides essential methods and preventive measures for ensuring pipeline integrity management and safe operation throughout the practical design and production management process of pipelines.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110606"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2026-01-19DOI: 10.1016/j.engfailanal.2026.110595
Masoud Behzad , Mehdi Behzad , Somaye Mohammadi , Mohammad Erfan Yadegari , Mohammad Haghighi
Electro-pumps are critical assets in water distribution networks, yet their performance is compromised by scaling phenomena. Given the interdependence of pump maintenance and associated components such as the motor, coupling, and structure, an integrated framework encompassing all equipment components is needed. This paper presents a maintenance framework for electro-pumping systems that supports maintenance planning by converting vibration and thermographic condition monitoring results into prioritized maintenance actions. Faults are identified and ranked based on their severity and frequency of occurrence, providing practical input for maintenance scheduling and resource allocation. For vibration measurements, sensors were mounted on the drive and non-drive ends of both the motors and pumps. These measurements cover low (10–1,000 Hz) and high (1,000–8,000 Hz) frequency ranges to ensure adequate fault detection resolution. Thermographic images of the pumps were captured in the final round following vibration measurements. The thermographic findings were in strong agreement with the vibration analysis, confirming coupling-related anomalies. The maintenance actions with the highest priority scores were inspecting bearing lubrication and replacement (1.35), resolving coupling issues (1.13), and performing alignment (0.91). In the ranking of maintenance priorities derived from the prioritized maintenance actions, at least two pumps were classified at the highest priority level. The proposed framework not only prioritizes maintenance actions but also employs the derived maintenance priority index to identify critical pumps distributed across different water infrastructure systems on a regional scale. This approach supports failure prevention by enabling timely maintenance scheduling.
{"title":"A framework to prevent failure by mapping faults to maintenance actions in electro-pumps of water distribution networks","authors":"Masoud Behzad , Mehdi Behzad , Somaye Mohammadi , Mohammad Erfan Yadegari , Mohammad Haghighi","doi":"10.1016/j.engfailanal.2026.110595","DOIUrl":"10.1016/j.engfailanal.2026.110595","url":null,"abstract":"<div><div>Electro-pumps are critical assets in water distribution networks, yet their performance is compromised by scaling phenomena. Given the interdependence of pump maintenance and associated components such as the motor, coupling, and structure, an integrated framework encompassing all equipment components is needed. This paper presents a maintenance framework for electro-pumping systems that supports maintenance planning by converting vibration and thermographic condition monitoring results into prioritized maintenance actions. Faults are identified and ranked based on their severity and frequency of occurrence, providing practical input for maintenance scheduling and resource allocation. For vibration measurements, sensors were mounted on the drive and non-drive ends of both the motors and pumps. These measurements cover low (10–1,000 Hz) and high (1,000–8,000 Hz) frequency ranges to ensure adequate fault detection resolution. Thermographic images of the pumps were captured in the final round following vibration measurements. The thermographic findings were in strong agreement with the vibration analysis, confirming coupling-related anomalies. The maintenance actions with the highest priority scores were inspecting bearing lubrication and replacement (1.35), resolving coupling issues (1.13), and performing alignment (0.91). In the ranking of maintenance priorities derived from the prioritized maintenance actions, at least two pumps were classified at the highest priority level. The proposed framework not only prioritizes maintenance actions but also employs the derived maintenance priority index to identify critical pumps distributed across different water infrastructure systems on a regional scale. This approach supports failure prevention by enabling timely maintenance scheduling.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110595"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075805","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2026-01-16DOI: 10.1016/j.engfailanal.2026.110575
Flavia da Cruz Gallo , Mariel Ojeda-Tuz , Ciaran O’Rourke , Ryan Catarelli , Jennifer Bridge
Multiple aluminum light poles in Florida failed during recent hurricanes at wind speeds below design specifications, raising concerns about structural integrity and manufacturing quality. This study investigates the failure of cast A356-T6 aluminum alloy bases through combined metallurgical analysis and finite element modeling. Computed Tomography (CT) and metallography revealed critical porosity levels (5.9 – 8.2 %) exceeding ASTM acceptance thresholds, along with microstructural variability near bolt holes. Local hardness and tensile testing indicated reduced yield strength, approximately 20 – 25 % relative to nominal A356-T6 values, consistent with casting defects and installation-induced pre-strain, significantly narrowing the safety margin under hurricane winds. Fracture morphology confirmed monotonic overload rather than fatigue. A full-scale finite element model of the Lake Jessup Bridge base assembly was developed to evaluate stress distribution under design-level and hurricane-level wind loading while incorporating measured material properties and installation irregularities. Simulations showed that when porosity-reduced strength was combined with geometric stress risers and uneven leveling-nut preload, localized stresses exceeded the experimentally measured yield strength (∼116 MPa) even under wind speeds below design thresholds. This study is the first to integrate CT-quantified porosity, field installation audits, tensile testing, and wind-driven structural modeling to explain premature hurricane-induced failures of cast aluminum pole bases. The findings demonstrate that premature failures resulted from the synergistic interaction of casting defects, geometric vulnerability, and installation-induced overloads rather than a single governing mechanism. Recommendations include stricter casting quality control, torque-limiting installation protocols, and minor design modifications aimed at reducing stress concentrations and improving reliability of aluminum infrastructure in hurricane-prone regions.
{"title":"Failure analysis of cast aluminum A356-T6 light pole bases following Catastrophic Hurricane Exposure: Microstructural, Mechanical, and Fractographic investigations","authors":"Flavia da Cruz Gallo , Mariel Ojeda-Tuz , Ciaran O’Rourke , Ryan Catarelli , Jennifer Bridge","doi":"10.1016/j.engfailanal.2026.110575","DOIUrl":"10.1016/j.engfailanal.2026.110575","url":null,"abstract":"<div><div>Multiple aluminum light poles in Florida failed during recent hurricanes at wind speeds below design specifications, raising concerns about structural integrity and manufacturing quality. This study investigates the failure of cast A356-T6 aluminum alloy bases through combined metallurgical analysis and finite element modeling. Computed Tomography (CT) and metallography revealed critical porosity levels (5.9 – 8.2 %) exceeding ASTM acceptance thresholds, along with microstructural variability near bolt holes. Local hardness and tensile testing indicated reduced yield strength, approximately 20 – 25 % relative to nominal A356-T6 values, consistent with casting defects and installation-induced pre-strain, significantly narrowing the safety margin under hurricane winds. Fracture morphology confirmed monotonic overload rather than fatigue. A full-scale finite element model of the Lake Jessup Bridge base assembly was developed to evaluate stress distribution under design-level and hurricane-level wind loading while incorporating measured material properties and installation irregularities. Simulations showed that when porosity-reduced strength was combined with geometric stress risers and uneven leveling-nut preload, localized stresses exceeded the experimentally measured yield strength (∼116 MPa) even under wind speeds below design thresholds. This study is the first to integrate CT-quantified porosity, field installation audits, tensile testing, and wind-driven structural modeling to explain premature hurricane-induced failures of cast aluminum pole bases. The findings demonstrate that premature failures resulted from the synergistic interaction of casting defects, geometric vulnerability, and installation-induced overloads rather than a single governing mechanism. Recommendations include stricter casting quality control, torque-limiting installation protocols, and minor design modifications aimed at reducing stress concentrations and improving reliability of aluminum infrastructure in hurricane-prone regions.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110575"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146001771","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2026-01-19DOI: 10.1016/j.engfailanal.2026.110593
MN James , D Bernard , C Louis , L. Matthews , DG Hattingh
This failure investigation deals with an apparently straightforward case of gear failure that was undertaken to see whether it contained any cracking that pre-existed the final fracture of a tooth. Some interesting features were observed in this case, as there were several cracks in the gear besides the primary one that led to fracture, showing evidence of at least two different crack initiation mechanisms. These mechanisms were intergranular quench cracking, occasionally involving large inclusions where chevron markings were seen on the ductile fracture surface that pointed back to the local fracture origin, and the observation of very localised intergranular oxidation in the carburised case of the gear that occurred during heat treatment. Both mechanisms of crack initiation then led to fatigue cracking. The final conclusion was that the gears failed prematurely from fatigue cracking initiated by intergranular oxidation combined with an impact load.
{"title":"Failure investigation of gears on a tyre manufacturing machine","authors":"MN James , D Bernard , C Louis , L. Matthews , DG Hattingh","doi":"10.1016/j.engfailanal.2026.110593","DOIUrl":"10.1016/j.engfailanal.2026.110593","url":null,"abstract":"<div><div>This failure investigation deals with an apparently straightforward case of gear failure that was undertaken to see whether it contained any cracking that pre-existed the final fracture of a tooth. Some interesting features were observed in this case, as there were several cracks in the gear besides the primary one that led to fracture, showing evidence of at least two different crack initiation mechanisms. These mechanisms were intergranular quench cracking, occasionally involving large inclusions where chevron markings were seen on the ductile fracture surface that pointed back to the local fracture origin, and the observation of very localised intergranular oxidation in the carburised case of the gear that occurred during heat treatment. Both mechanisms of crack initiation then led to fatigue cracking. The final conclusion was that the gears failed prematurely from fatigue cracking initiated by intergranular oxidation combined with an impact load.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110593"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146025777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2026-01-14DOI: 10.1016/j.engfailanal.2026.110581
Pengfei Ma , Yangyang Zhang , Lichao Nie , Zhiqiang Li , Zhicheng Song , Yuancheng Li
Collapse is likely when long, deeply buried tunnels intersect water-rich fractured basalt. Single-method forecasting cannot resolve the spatial distribution of water-bearing structures or the progressive failure process. Using the Xianglu Mountain Tunnel as a case study, we propose a mechanism-analysis framework in which integrated geophysical prospecting provides priors for a Peridynamic (PD) numerical model. Specifically, Seismic ahead prospecting yields spatial distributions of elastic modulus, Poisson’s ratio, and density ahead of the face. Direct current resistivity delineates low-resistivity anomalies and, through an empirical resistivity–permeability relationship, enables quantitative inversion of the pre-excavation seepage field. These geophysical products are then injected as prior fields into a PD-based excavation–seepage failure model. The simulations indicate progressive damage of the confining rock layer (aquiclude) under multi-factor coupling until the damage zone coalesces and collapse occurs. Comparison with field observations shows close agreement in the predicted affected extent, demonstrating that the integrated approach explains collapse during excavation in water-rich basalt tunnels and provides a reliable pathway for advanced prevention and control of similar geohazards in deeply buried tunnels.
{"title":"Collapse mechanism of deep-buried long water-rich basalt tunnels based on integrated geophysical prospecting: A case study in Yunnan, China","authors":"Pengfei Ma , Yangyang Zhang , Lichao Nie , Zhiqiang Li , Zhicheng Song , Yuancheng Li","doi":"10.1016/j.engfailanal.2026.110581","DOIUrl":"10.1016/j.engfailanal.2026.110581","url":null,"abstract":"<div><div>Collapse is likely when long, deeply buried tunnels intersect water-rich fractured basalt. Single-method forecasting cannot resolve the spatial distribution of water-bearing structures or the progressive failure process. Using the Xianglu Mountain Tunnel as a case study, we propose a mechanism-analysis framework in which integrated geophysical prospecting provides priors for a Peridynamic (PD) numerical model. Specifically, Seismic ahead prospecting yields spatial distributions of elastic modulus, Poisson’s ratio, and density ahead of the face. Direct current resistivity delineates low-resistivity anomalies and, through an empirical resistivity–permeability relationship, enables quantitative inversion of the pre-excavation seepage field. These geophysical products are then injected as prior fields into a PD-based excavation–seepage failure model. The simulations indicate progressive damage of the confining rock layer (aquiclude) under multi-factor coupling until the damage zone coalesces and collapse occurs. Comparison with field observations shows close agreement in the predicted affected extent, demonstrating that the integrated approach explains collapse during excavation in water-rich basalt tunnels and provides a reliable pathway for advanced prevention and control of similar geohazards in deeply buried tunnels.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110581"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146025776","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Electrical discharge machining (EDM) is widely used in the manufacturing of H13 die steel due to its capability for processing complex geometries and high-hardness materials. Despite its advantages, EDM introduces surface defects—such as recast layers, microcracks, and residual tensile stresses—that significantly promote early-stage crack initiation and accelerate crack propagation under thermal cycling. This study investigates the damage accumulation behavior of H13 hot-work die steel subjected to EDM and evaluates the crack suppression potential of shot peening (SP) at different Almen intensities (0.11–0.28 mmA). Thermal cycling tests were performed over 450 to 1500 cycles (with a step size of 150 cycles) using a self-constrained thermal cycling system. Based on the Uddeholm standard, each thermal cycle involved heating from room temperature to 700 °C within 5.5 s, followed by a 17 s cooling phase. Compared to ground specimens, EDM-processed samples exhibited earlier crack initiation, more complex branched crack networks, and deeper main cracks, with a maximum depth of 684.2 μm. SP was subsequently applied to induce stress and hardness gradients in surface-near regions, effectively reshaping the mechanical conditions that control crack evolution. The results reveal a nonlinear correlation between SP intensity and mitigation of cracking under thermal cycling. While moderate intensities may induce stress reversals or sub-surface microstructural instabilities, the optimal SP intensity (0.28 mmA) generated a compressive residual stress field (∼–700 MPa, 150 μm depth) and a gradual hardness gradient (∼180 μm), forming an effective barrier to crack extension. These surface integrity gradients significantly delayed crack coalescence and reduced overall damage accumulation. This study highlights that intensity-optimized SP is an effective strategy for altering the crack driving force distribution in EDM-affected regions and for improving the structural endurance of hot-work tooling under cyclic thermal loading.
{"title":"Enhancement of resistance to cracking under thermal cycling of EDM-treated H13 steel by shot peening with optimized intensity","authors":"Pengpeng Zuo , Zhiyang Dou , Huikai Yang , Haoyan Hou , Yafeng Zheng","doi":"10.1016/j.engfailanal.2026.110622","DOIUrl":"10.1016/j.engfailanal.2026.110622","url":null,"abstract":"<div><div>Electrical discharge machining (EDM) is widely used in the manufacturing of H13 die steel due to its capability for processing complex geometries and high-hardness materials. Despite its advantages, EDM introduces surface defects—such as recast layers, microcracks, and residual tensile stresses—that significantly promote early-stage crack initiation and accelerate crack propagation under thermal cycling. This study investigates the damage accumulation behavior of H13 hot-work die steel subjected to EDM and evaluates the crack suppression potential of shot peening (SP) at different Almen intensities (0.11–0.28 mmA). Thermal cycling tests were performed over 450 to 1500 cycles (with a step size of 150 cycles) using a self-constrained thermal cycling system. Based on the Uddeholm standard, each thermal cycle involved heating from room temperature to 700 °C within 5.5 s, followed by a 17 s cooling phase. Compared to ground specimens, EDM-processed samples exhibited earlier crack initiation, more complex branched crack networks, and deeper main cracks, with a maximum depth of 684.2 μm. SP was subsequently applied to induce stress and hardness gradients in surface-near regions, effectively reshaping the mechanical conditions that control crack evolution. The results reveal a nonlinear correlation between SP intensity and mitigation of cracking under thermal cycling. While moderate intensities may induce stress reversals or sub-surface microstructural instabilities, the optimal SP intensity (0.28 mmA) generated a compressive residual stress field (∼–700 MPa, 150 μm depth) and a gradual hardness gradient (∼180 μm), forming an effective barrier to crack extension. These surface integrity gradients significantly delayed crack coalescence and reduced overall damage accumulation. This study highlights that intensity-optimized SP is an effective strategy for altering the crack driving force distribution in EDM-affected regions and for improving the structural endurance of hot-work tooling under cyclic thermal loading.</div></div>","PeriodicalId":11677,"journal":{"name":"Engineering Failure Analysis","volume":"187 ","pages":"Article 110622"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075745","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号