Engineering Structures最新文献_第6页

Identification of mechanism-specific persistent scatterers for enhancing MT-InSAR in Structural Health Monitoring applications 在结构健康监测应用中增强MT-InSAR的机制特异性持久性散射体的识别

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-08 DOI: 10.1016/j.engstruct.2026.122103

Hao Kuai , Valentina Macchiarulo , Satyadhrik Sharma , Pantelis Karamitopoulos , Francesco Messali , Alice Cicirello , Giorgia Giardina

Growing incidents of structural damage and failures underscore the urgent need for more advanced Structural Health Monitoring (SHM) solutions. While Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR) has revolutionised SHM by enabling automated, long-term, and large-scale displacement monitoring of structures using Persistent Scatterers (PSs), its applicability is often constrained by the unpredictable spatial distribution of PSs. Conventional suitability assessments that rely primarily on PS density fail to account for the underlying structural behaviours, limiting their reliability.

This paper introduces a novel structural-based inverse approach that uniquely integrates MT-InSAR characteristics with structural response modelling to overcome these limitations. Unlike existing approaches, the method explicitly evaluates whether observed surface displacements adequately represent a target damage mechanism by comparing outputs from a pseudo sensor with those from a virtual MT-InSAR sensor. If this condition is satisfied, it then determines the minimum required number and optimal spatial arrangement of ideal PSs using modified pivoted QR factorisation, where satellite-induced positional uncertainties are rigorously modelled through Radial Basis Function kernels.

The proposed method was validated on a quay wall in Amsterdam using Finite Element Method (FEM) simulations of three distinct damage mechanisms. Results demonstrate its unique capability to quantitatively assess displacement representativeness and to pinpoint ideal PSs for robust monitoring. Leveraging these insights, the method was further applied to evaluate MT-InSAR monitoring feasibility across Amsterdam’s historic centre, successfully identifying quay wall segments amenable to reliable observation. This work represents a significant advancement in MT-InSAR-based SHM, providing a more targeted and structurally informed approach for real-world infrastructure monitoring.

越来越多的结构损坏和失效事件强调了对更先进的结构健康监测（SHM）解决方案的迫切需要。虽然多时相干涉合成孔径雷达（MT-InSAR）通过使用持续散射体（ps）对结构进行自动化、长期和大规模位移监测，彻底改变了SHM，但其适用性往往受到不可预测的ps空间分布的限制。传统的适用性评估主要依赖于PS密度，无法解释潜在的结构行为，限制了它们的可靠性。本文介绍了一种新的基于结构的反演方法，该方法独特地将MT-InSAR特征与结构响应建模相结合，以克服这些局限性。与现有方法不同，该方法通过比较伪传感器与虚拟MT-InSAR传感器的输出，明确评估观察到的表面位移是否充分代表目标损伤机制。如果满足此条件，则使用改进的枢轴QR分解确定理想PSs的最小所需数量和最佳空间安排，其中卫星引起的位置不确定性通过径向基函数核严格建模。采用三种不同损伤机制的有限元模拟，对阿姆斯特丹某码头墙体进行了验证。结果表明，其独特的能力，定量评估位移代表性和确定理想的ps稳健监测。利用这些见解，该方法被进一步应用于评估MT-InSAR监测阿姆斯特丹历史中心的可行性，成功地确定了适合可靠观察的码头墙段。这项工作代表了基于mt - insar的SHM的重大进步，为现实世界的基础设施监测提供了更有针对性和结构知情的方法。

{"title":"Identification of mechanism-specific persistent scatterers for enhancing MT-InSAR in Structural Health Monitoring applications","authors":"Hao Kuai , Valentina Macchiarulo , Satyadhrik Sharma , Pantelis Karamitopoulos , Francesco Messali , Alice Cicirello , Giorgia Giardina","doi":"10.1016/j.engstruct.2026.122103","DOIUrl":"10.1016/j.engstruct.2026.122103","url":null,"abstract":"<div><div>Growing incidents of structural damage and failures underscore the urgent need for more advanced Structural Health Monitoring (SHM) solutions. While Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR) has revolutionised SHM by enabling automated, long-term, and large-scale displacement monitoring of structures using Persistent Scatterers (PSs), its applicability is often constrained by the unpredictable spatial distribution of PSs. Conventional suitability assessments that rely primarily on PS density fail to account for the underlying structural behaviours, limiting their reliability.</div><div>This paper introduces a novel structural-based inverse approach that uniquely integrates MT-InSAR characteristics with structural response modelling to overcome these limitations. Unlike existing approaches, the method explicitly evaluates whether observed surface displacements adequately represent a target damage mechanism by comparing outputs from a pseudo sensor with those from a virtual MT-InSAR sensor. If this condition is satisfied, it then determines the minimum required number and optimal spatial arrangement of ideal PSs using modified pivoted QR factorisation, where satellite-induced positional uncertainties are rigorously modelled through Radial Basis Function kernels.</div><div>The proposed method was validated on a quay wall in Amsterdam using Finite Element Method (FEM) simulations of three distinct damage mechanisms. Results demonstrate its unique capability to quantitatively assess displacement representativeness and to pinpoint ideal PSs for robust monitoring. Leveraging these insights, the method was further applied to evaluate MT-InSAR monitoring feasibility across Amsterdam’s historic centre, successfully identifying quay wall segments amenable to reliable observation. This work represents a significant advancement in MT-InSAR-based SHM, providing a more targeted and structurally informed approach for real-world infrastructure monitoring.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122103"},"PeriodicalIF":6.4,"publicationDate":"2026-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940830","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

Global dynamics and response for heave mitigation of semi-submersible platform using vibration isolation system with customized nonlinear forces 采用自定义非线性力隔振系统的半潜式平台整体动力学及响应

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-07 DOI: 10.1016/j.engstruct.2025.122021

Yuxi Li , Zhouchao Wei , Lin Wang , Tomasz Kapitaniak

Excessive heave motion and harmful vibrations have a significant influence on the semi-submersible platform (SSP), which may lead to structural damage even capsizing or sinking. In this paper, by introducing roller-raceway-disc spring structure (RRDS), a new vibration isolation system (VIS) with customized nonlinear forces (CNFs) is proposed to mitigate the heave motion of SSP. According to the two different customized designs of the raceway in RRDS, two different types of common and important forces in research field of nonlinear vibration are customized, i.e., polynomial nonlinear force (PNNF) and piecewise nonlinear force (PWNF). Then, global dynamics and heave response of the SSP using this new VIS with two different CNFs are investigated through analytical and numerical methods. The parameter conditions of global bifurcation and chaotic dynamics are strictly obtained by classical Melnikov method and piecewise Melnikov method. Chaotic characteristics and regions of the SSP system’s parameters are shown detailedly. Moreover, the amplitude–frequency response and force transmissibility of the corresponding VIS about heave mitigation are analyzed by the averaging method. The influence of various SSP system’s parameters on the response and vibration isolation effect is comprehensively analyzed and demonstrated. In addition, the relevant theoretical results are verify and discussed using actual parameters of a class of SSPs through numerical simulations. These results demonstrate the rich dynamics and excellent vibration isolation effect brought by the proposed VIS with CNFs about heave mitigation.

过度的升沉运动和有害振动对半潜式平台有很大的影响，可能导致结构破坏甚至倾覆或沉没。本文通过引入滚子-滚道-盘式弹簧结构（RRDS），提出了一种具有自定义非线性力（CNFs）的新型隔振系统（VIS），以减轻SSP的升沉运动。针对RRDS中滚道的两种不同定制设计，定制了非线性振动研究领域中常见且重要的两种不同类型的力，即多项式非线性力（PNNF）和分段非线性力（PWNF）。在此基础上，通过解析和数值方法研究了两种不同CNFs下该系统的整体动力学和升沉响应。采用经典Melnikov方法和分段Melnikov方法，严格地得到了全局分岔和混沌动力学的参数条件。详细分析了SSP系统参数的混沌特性和混沌区域。此外，采用平均法分析了相应VIS的幅频响应和力传递率。综合分析和论证了SSP系统各参数对系统响应和隔振效果的影响。此外，通过数值模拟，利用一类ssp的实际参数对相关理论结果进行了验证和讨论。结果表明，该系统具有丰富的动力学特性和良好的隔振效果。

{"title":"Global dynamics and response for heave mitigation of semi-submersible platform using vibration isolation system with customized nonlinear forces","authors":"Yuxi Li , Zhouchao Wei , Lin Wang , Tomasz Kapitaniak","doi":"10.1016/j.engstruct.2025.122021","DOIUrl":"10.1016/j.engstruct.2025.122021","url":null,"abstract":"<div><div>Excessive heave motion and harmful vibrations have a significant influence on the semi-submersible platform (SSP), which may lead to structural damage even capsizing or sinking. In this paper, by introducing roller-raceway-disc spring structure (RRDS), a new vibration isolation system (VIS) with customized nonlinear forces (CNFs) is proposed to mitigate the heave motion of SSP. According to the two different customized designs of the raceway in RRDS, two different types of common and important forces in research field of nonlinear vibration are customized, i.e., polynomial nonlinear force (PNNF) and piecewise nonlinear force (PWNF). Then, global dynamics and heave response of the SSP using this new VIS with two different CNFs are investigated through analytical and numerical methods. The parameter conditions of global bifurcation and chaotic dynamics are strictly obtained by classical Melnikov method and piecewise Melnikov method. Chaotic characteristics and regions of the SSP system’s parameters are shown detailedly. Moreover, the amplitude–frequency response and force transmissibility of the corresponding VIS about heave mitigation are analyzed by the averaging method. The influence of various SSP system’s parameters on the response and vibration isolation effect is comprehensively analyzed and demonstrated. In addition, the relevant theoretical results are verify and discussed using actual parameters of a class of SSPs through numerical simulations. These results demonstrate the rich dynamics and excellent vibration isolation effect brought by the proposed VIS with CNFs about heave mitigation.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122021"},"PeriodicalIF":6.4,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940959","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 modal performance-driven topology optimization framework for structural vibration avoidance 一种模态性能驱动的结构避振拓扑优化框架

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-07 DOI: 10.1016/j.engstruct.2026.122094

Kai Huang , Zhengguang Li , Shitong Yang

Structural topology optimization enables vibration attenuation through relocation of natural frequencies beyond excitation bands and the efficacy of this avoidance strategy degrades considerably with increasing excitation bandwidth. In contrast, by optimizing the mode shapes of the structure to align their nodes with critical locations (such as the rocket payload or optical benches), vibration at these positions can be significantly reduced. Thus, in this paper, a novel topology optimization framework targeting modal dynamic performance is proposed for structural vibration avoidance. Within this framework, the modal response characteristics are incorporated into the objective function, and the Solid Isotropic Material with Penalization (SIMP) method is employed for topology optimization. The proposed approach provides an effective solution to the key challenges encountered in modal topology optimization, including mode tracking, mode switching, sensitivity analysis of both distinct and repeated modes, and avoidance of artificial modes. Several numerical examples are given to demonstrate the effectiveness of the proposed method.

结构拓扑优化可以通过将固有频率重新定位到激励频带之外来实现振动衰减，但随着激励频带的增加，这种避免策略的效果会显著降低。相比之下，通过优化结构的模态振型，使节点与关键位置（如火箭有效载荷或光学工作台）对齐，可以显著减少这些位置的振动。为此，本文提出了一种以模态动力性能为目标的新型结构避振拓扑优化框架。在此框架内，将模态响应特性纳入目标函数，并采用固体各向同性材料惩罚法（SIMP）进行拓扑优化。该方法有效地解决了模态拓扑优化中遇到的关键问题，包括模态跟踪、模态切换、不同模态和重复模态的灵敏度分析以及人为模态的避免。算例验证了该方法的有效性。

{"title":"A modal performance-driven topology optimization framework for structural vibration avoidance","authors":"Kai Huang , Zhengguang Li , Shitong Yang","doi":"10.1016/j.engstruct.2026.122094","DOIUrl":"10.1016/j.engstruct.2026.122094","url":null,"abstract":"<div><div>Structural topology optimization enables vibration attenuation through relocation of natural frequencies beyond excitation bands and the efficacy of this avoidance strategy degrades considerably with increasing excitation bandwidth. In contrast, by optimizing the mode shapes of the structure to align their nodes with critical locations (such as the rocket payload or optical benches), vibration at these positions can be significantly reduced. Thus, in this paper, a novel topology optimization framework targeting modal dynamic performance is proposed for structural vibration avoidance. Within this framework, the modal response characteristics are incorporated into the objective function, and the Solid Isotropic Material with Penalization (SIMP) method is employed for topology optimization. The proposed approach provides an effective solution to the key challenges encountered in modal topology optimization, including mode tracking, mode switching, sensitivity analysis of both distinct and repeated modes, and avoidance of artificial modes. Several numerical examples are given to demonstrate the effectiveness of the proposed method.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122094"},"PeriodicalIF":6.4,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940887","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

Prediction of seismic capacity limit states of corroded bridge piers considering climate change: A multi-task learning approach 考虑气候变化的腐蚀桥墩抗震能力极限状态预测：一种多任务学习方法

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-07 DOI: 10.1016/j.engstruct.2025.122081

Pouya Moeinifard, Mamdouh El-Badry

Seismic capacity limit states (LSs) of reinforced concrete (RC) bridges deteriorate over time due to corrosion, a process further exacerbated by climate change. This compounding effect necessitates a time-dependent, multi-hazard capacity model. However, the inherent complexity and uncertainty associated with compounding effects of corrosion and climate change, make conventional methods computationally intensive, particularly in probabilistic seismic risk assessments involving large bridge inventories. This study proposes a novel Multi-Task Learning (MTL)-based machine learning framework that enables different tasks to share information for improved prediction of capacity LSs and failure modes of corroded bridge piers. A comprehensive synthetic dataset representing typical highway bridges under different environmental scenarios is generated to train the proposed MTL model and several conventional single-task learning (STL) models. All models are optimized by hyperparameter tuning, and their results were subsequently interpreted to assess the contribution of key input variables. The results show that the MTL framework outperforms the STL models in terms of accuracy, generalization, and computational efficiency. The proposed model is then integrated into a fragility framework to predict LSs. Based on these predictions, the fragility curves are generated and compared with those obtained from finite element analysis for the case studies, showing close agreement and demonstrating the reliability of the proposed method. The framework provides a rapid assessment tool for evaluating the seismic capacity of aging, corroded bridges under climate change conditions, offering utility in both design and maintenance applications.

钢筋混凝土（RC）桥梁的抗震能力极限状态（LSs）由于腐蚀而随着时间的推移而恶化，气候变化进一步加剧了这一过程。这种复合效应需要一个随时间变化的多灾害能力模型。然而，固有的复杂性和不确定性与腐蚀和气候变化的复合效应有关，使得传统方法的计算量很大，特别是在涉及大型桥梁库存的概率地震风险评估中。本研究提出了一种新的基于多任务学习（MTL）的机器学习框架，该框架使不同的任务能够共享信息，以改进对腐蚀桥墩的容量LSs和失效模式的预测。生成了代表不同环境情景下典型公路桥梁的综合合成数据集，用于训练所提出的MTL模型和几种传统的单任务学习（STL）模型。所有模型都通过超参数调整进行优化，其结果随后被解释以评估关键输入变量的贡献。结果表明，MTL框架在精度、泛化和计算效率方面优于STL模型。然后将提出的模型集成到脆弱性框架中以预测LSs。基于这些预测，生成了脆性曲线，并将其与有限元分析所得的脆性曲线进行了比较，结果显示出非常接近的一致性，并证明了所提出方法的可靠性。该框架为评估气候变化条件下老化、腐蚀桥梁的抗震能力提供了快速评估工具，在设计和维护应用中都具有实用性。

{"title":"Prediction of seismic capacity limit states of corroded bridge piers considering climate change: A multi-task learning approach","authors":"Pouya Moeinifard, Mamdouh El-Badry","doi":"10.1016/j.engstruct.2025.122081","DOIUrl":"10.1016/j.engstruct.2025.122081","url":null,"abstract":"<div><div>Seismic capacity limit states (LSs) of reinforced concrete (RC) bridges deteriorate over time due to corrosion, a process further exacerbated by climate change. This compounding effect necessitates a time-dependent, multi-hazard capacity model. However, the inherent complexity and uncertainty associated with compounding effects of corrosion and climate change, make conventional methods computationally intensive, particularly in probabilistic seismic risk assessments involving large bridge inventories. This study proposes a novel Multi-Task Learning (MTL)-based machine learning framework that enables different tasks to share information for improved prediction of capacity LSs and failure modes of corroded bridge piers. A comprehensive synthetic dataset representing typical highway bridges under different environmental scenarios is generated to train the proposed MTL model and several conventional single-task learning (STL) models. All models are optimized by hyperparameter tuning, and their results were subsequently interpreted to assess the contribution of key input variables. The results show that the MTL framework outperforms the STL models in terms of accuracy, generalization, and computational efficiency. The proposed model is then integrated into a fragility framework to predict LSs. Based on these predictions, the fragility curves are generated and compared with those obtained from finite element analysis for the case studies, showing close agreement and demonstrating the reliability of the proposed method. The framework provides a rapid assessment tool for evaluating the seismic capacity of aging, corroded bridges under climate change conditions, offering utility in both design and maintenance applications.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122081"},"PeriodicalIF":6.4,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940889","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

Structural design of straddle-type composite guideways based on NSGA-II optimization: Performance enhancement and experimental validation 基于NSGA-II优化的跨座式复合导轨结构设计：性能提升与实验验证

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-07 DOI: 10.1016/j.engstruct.2025.122076

Zhengwei Bai, Eryu Zhu, Wenchao Cai, Honghe Jian, Yiming Li

This study proposed a multi-objective optimisation method based on the second-generation non-dominated sorting genetic algorithm (NSGA-II) for the design of straddle-type composite guideways, aiming to enhance economic efficiency and mechanical performance. Through optimisation, solutions focused on economic efficiency, steel use, and carbon emissions were identified, and section parameters were simplified for experimental feasibility. Experimental models with an original span of 25 m and a 10:1 scale ratio (actual span of 2.5 m) were constructed to validate the results. The optimised designs (FN01 and FN02) showed improvements in ultimate load capacity of 13.26 % and 16.48 %, respectively, compared with the original scheme (FN00) in finite element analysis, with ultimate vertical deformation increasing by 0.92 % and 9.94 %, respectively. Experimental results confirmed that the optimised guideways (FN01 and FN02) increased the ultimate load capacity by 42.57 % and 49.37 %, respectively, with vertical deformation rising by 23.13 % and 22.35 %, respectively, compared with FN00. Section optimisation enhanced the plastic deformation capacity, indicating improved load-bearing capacity, ductility, and toughness. This study provides a theoretical framework for designing straddle-type composite guideways, with significant potential to improve structural performance, reduce material consumption, and lower carbon emissions.

本研究提出了一种基于第二代非主导排序遗传算法（NSGA-II）的跨座式复合导轨多目标优化设计方法，旨在提高跨座式复合导轨的经济效益和力学性能。通过优化，确定了以经济效率、钢材使用和碳排放为重点的解决方案，并简化了截面参数以实现实验可行性。建立了原始跨度为25 m，比例尺为10:1（实际跨度为2.5 m）的实验模型来验证结果。优化方案（FN01和FN02）的极限承载力比原方案（FN00）分别提高了13.26 %和16.48 %，竖向极限变形分别提高了0.92 %和9.94 %。试验结果表明，优化后的导轨（FN01和FN02）的极限承载能力分别比FN00提高42.57 %和49.37 %，竖向变形分别比FN00提高23.13 %和22.35 %。截面优化增强了塑性变形能力，表明承载能力、延性和韧性得到改善。该研究为跨座式复合导轨的设计提供了理论框架，在提高结构性能、减少材料消耗和降低碳排放方面具有重要的潜力。

{"title":"Structural design of straddle-type composite guideways based on NSGA-II optimization: Performance enhancement and experimental validation","authors":"Zhengwei Bai, Eryu Zhu, Wenchao Cai, Honghe Jian, Yiming Li","doi":"10.1016/j.engstruct.2025.122076","DOIUrl":"10.1016/j.engstruct.2025.122076","url":null,"abstract":"<div><div>This study proposed a multi-objective optimisation method based on the second-generation non-dominated sorting genetic algorithm (NSGA-II) for the design of straddle-type composite guideways, aiming to enhance economic efficiency and mechanical performance. Through optimisation, solutions focused on economic efficiency, steel use, and carbon emissions were identified, and section parameters were simplified for experimental feasibility. Experimental models with an original span of 25 m and a 10:1 scale ratio (actual span of 2.5 m) were constructed to validate the results. The optimised designs (FN01 and FN02) showed improvements in ultimate load capacity of 13.26 % and 16.48 %, respectively, compared with the original scheme (FN00) in finite element analysis, with ultimate vertical deformation increasing by 0.92 % and 9.94 %, respectively. Experimental results confirmed that the optimised guideways (FN01 and FN02) increased the ultimate load capacity by 42.57 % and 49.37 %, respectively, with vertical deformation rising by 23.13 % and 22.35 %, respectively, compared with FN00. Section optimisation enhanced the plastic deformation capacity, indicating improved load-bearing capacity, ductility, and toughness. This study provides a theoretical framework for designing straddle-type composite guideways, with significant potential to improve structural performance, reduce material consumption, and lower carbon emissions.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122076"},"PeriodicalIF":6.4,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940958","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

Region-based digital image correlation framework to monitor UHPC girders under cyclic loads using transformer-based segmentation 基于变压器分割的区域数字图像相关框架监测循环荷载作用下的超高压混凝土梁

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-07 DOI: 10.1016/j.engstruct.2026.122102

Mostafa Iraniparast, Peng Patrick Sun, Georgios Apostolakis, Kevin R. Mackie

Reversed cyclic testing to failure of large-scale reinforced concrete members poses significant challenges in tracking strains, curvatures, cracks, and plastic hinge formation. Traditional instrumentation provides only pointwise measurements, while photogrammetry-based methods such as conventional digital image correlation (DIC) struggle with discontinuities caused by surface damage. To enhance strain measurement near cracks, this study proposes an automated damage assessment method utilizing a multi-vision setup and a region-based DIC framework enhanced by vision transformer (ViT)-based segmentation. The proposed vision methodology also derives curvature profiles from fused displacement data to monitor plastic hinge evolution and automate damage evaluation of full-scale ultra-high-performance concrete (UHPC) girders under cyclic loading. UHPC is an emerging structural material known for its exceptional strength, ductility, durability, and distributed cracking properties. Six large-scale UHPC girders were tested using conventional and lap-spliced reinforcement in cyclic four-point bending. One of the tested specimens was selected to validate the proposed methodology. Displacement measurements from both contact-based sensors and the non-contact multi-vision system demonstrated strong agreement. Region-based DIC yielded more accurate strain measurements near major cracks compared to conventional DIC. Curvature profiles and associated plastic-hinge rotations, which increased from about 0.012 rad (at the 1st peak of cycle 7 with 1.4 % drift) to about 0.049 rad (at the 2nd valley of cycle 10 with 2.4 % drift), and together with distinct jumps in the curvature deviation indices at the plastic-hinge location, effectively quantified cyclic nonlinearity, including peak responses. At the largest deformation level (cycle 10 with 2.4 % drift), the plastic-hinge rotations estimated by region-based DIC differed by about 3.0 % at the peak and 1.1 % at the valley relative to those obtained using AprilTag-based hinge kinematics, indicating good agreement between the two rotation estimation methods. In addition, image-based crack indicators captured and tracked the onset and progression of nonlinear behavior, with average crack width changed from 3.0 to 19.0 mm, crack area from 256 to 5297 mm², and crack depth ratio from 0.3 to 1.0.

大型钢筋混凝土构件失效的反向循环试验在跟踪应变、曲率、裂缝和塑性铰形成方面提出了重大挑战。传统仪器只能提供逐点测量，而基于摄影测量的方法，如传统的数字图像相关（DIC），则难以处理由表面损伤引起的不连续性。为了加强裂纹附近应变测量，本研究提出了一种基于多视觉设置和基于视觉变压器（ViT）分割增强的基于区域的DIC框架的自动损伤评估方法。提出的视觉方法还从融合位移数据中提取曲率轮廓，以监测塑性铰的演变，并自动评估循环荷载下全尺寸超高性能混凝土（UHPC）梁的损伤。UHPC是一种新兴的结构材料，以其卓越的强度、延展性、耐久性和分布开裂特性而闻名。对6根大型UHPC主梁进行了常规配筋和搭接配筋循环四点弯曲试验。选择一个测试样本来验证所提出的方法。来自接触式传感器和非接触式多视觉系统的位移测量结果显示出强烈的一致性。与传统DIC相比，基于区域的DIC在主要裂缝附近产生了更精确的应变测量。曲率曲线和相关的塑性铰旋转，从约0.012 rad（在第7个周期的第一个峰值，漂移1.4 %）增加到约0.049 rad（在第10个周期的第二个谷，漂移2.4 %），以及塑性铰位置曲率偏差指数的明显跳跃，有效地量化了循环非线性，包括峰值响应。在最大的变形水平（周期10，漂移2.4 %），相对于基于apriltag的铰链运动学，基于区域的DIC估计的塑性铰链旋转在峰值处相差约3.0 %，在谷处相差1.1 %，表明两种旋转估计方法之间具有良好的一致性。此外，基于图像的裂缝指标捕捉并跟踪了非线性行为的开始和发展，平均裂缝宽度从3.0到19.0 mm，裂缝面积从256到5297 mm²，裂缝深度比从0.3到1.0。

{"title":"Region-based digital image correlation framework to monitor UHPC girders under cyclic loads using transformer-based segmentation","authors":"Mostafa Iraniparast, Peng Patrick Sun, Georgios Apostolakis, Kevin R. Mackie","doi":"10.1016/j.engstruct.2026.122102","DOIUrl":"10.1016/j.engstruct.2026.122102","url":null,"abstract":"<div><div>Reversed cyclic testing to failure of large-scale reinforced concrete members poses significant challenges in tracking strains, curvatures, cracks, and plastic hinge formation. Traditional instrumentation provides only pointwise measurements, while photogrammetry-based methods such as conventional digital image correlation (DIC) struggle with discontinuities caused by surface damage. To enhance strain measurement near cracks, this study proposes an automated damage assessment method utilizing a multi-vision setup and a region-based DIC framework enhanced by vision transformer (ViT)-based segmentation. The proposed vision methodology also derives curvature profiles from fused displacement data to monitor plastic hinge evolution and automate damage evaluation of full-scale ultra-high-performance concrete (UHPC) girders under cyclic loading. UHPC is an emerging structural material known for its exceptional strength, ductility, durability, and distributed cracking properties. Six large-scale UHPC girders were tested using conventional and lap-spliced reinforcement in cyclic four-point bending. One of the tested specimens was selected to validate the proposed methodology. Displacement measurements from both contact-based sensors and the non-contact multi-vision system demonstrated strong agreement. Region-based DIC yielded more accurate strain measurements near major cracks compared to conventional DIC. Curvature profiles and associated plastic-hinge rotations, which increased from about 0.012 rad (at the 1st peak of cycle 7 with 1.4 % drift) to about 0.049 rad (at the 2nd valley of cycle 10 with 2.4 % drift), and together with distinct jumps in the curvature deviation indices at the plastic-hinge location, effectively quantified cyclic nonlinearity, including peak responses. At the largest deformation level (cycle 10 with 2.4 % drift), the plastic-hinge rotations estimated by region-based DIC differed by about 3.0 % at the peak and 1.1 % at the valley relative to those obtained using AprilTag-based hinge kinematics, indicating good agreement between the two rotation estimation methods. In addition, image-based crack indicators captured and tracked the onset and progression of nonlinear behavior, with average crack width changed from 3.0 to 19.0 mm, crack area from 256 to 5297 mm², and crack depth ratio from 0.3 to 1.0.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122102"},"PeriodicalIF":6.4,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940888","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

Influence of gypsum partitions on the damping ratio of steel moment frame buildings subjected to earthquakes 石膏隔墙对地震作用下钢弯矩框架结构阻尼比的影响

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-07 DOI: 10.1016/j.engstruct.2026.122114

Rosario Jouvhomme , Cristian Cruz

This paper quantifies the variation in modal damping ratios due to the deterioration of gypsum partitions and analyzes their dependence on the amplitude of structural response during earthquakes. A 9-story building with steel moment-resisting frames (SMRF) is modeled in OpenSees, explicitly incorporating gypsum partitions. The steel structure is modeled using linear-elastic elements, while partitions are modeled to account for nonlinear hysteretic behavior. Consequently, the only additional contribution to inherent damping comes from energy dissipation due to partition degradation. A multiple-stripe incremental analysis is conducted to obtain the structural response at different seismic intensities. For each computed response, system identification is employed to determine the first-mode damping ratio and fundamental period of an equivalent linear 9-story structure. The variation in the identified equivalent period and damping ratio with the amplitude of the structural response is then analyzed. A parametric study is conducted to assess the sensitivity of the identified results to: (1) the parameters defining the force-deformation relationship of the partitions; (2) the partition density per floor; (3) the type of partition employed; and (4) the amount of inherent damping in the structure. Finally, three additional SMRF buildings, consisting of 3, 20, and 50 stories are studied to evaluate the influence of building height on the inferred damping ratios. The analysis reveals that the damping provided by the partition walls is amplitude dependent. As amplitude increases, the contribution to damping reaches a plateau, and beyond a certain threshold, it begins to decrease. The partitions were found to increase the inherent damping ratio of the structures by a maximum of 8–67 %. In all cases, the damping ratios at amplitudes corresponding to the service-level earthquake scenario were higher than those at the maximum considered earthquake level. The results of this study offer valuable insights into the behavior of damping ratios during earthquakes, which can inform the development guidelines for specifying damping ratios in the seismic design of buildings.

本文量化了石膏隔层劣化引起的模态阻尼比的变化，并分析了它们与地震时结构反应幅度的关系。在OpenSees中，一个带有钢抗矩框架（SMRF）的9层建筑被建模，明确地结合了石膏分区。钢结构采用线弹性单元建模，而分区的建模是为了考虑非线性滞后行为。因此，唯一额外的贡献固有阻尼来自能量耗散由于分区退化。采用多条纹增量分析方法，得到了结构在不同烈度下的响应。对于每个计算得到的响应，采用系统识别来确定等效线性9层结构的一模态阻尼比和基本周期。分析了等效周期和阻尼比随结构响应幅值的变化规律。进行了参数化研究，以评估识别结果对：(1)定义分区力-变形关系的参数的敏感性；(2)每层隔断密度；（三）采用的隔断类型；(4)结构的固有阻尼量。最后，研究了另外三座SMRF建筑，包括3层、20层和50层，以评估建筑高度对推断阻尼比的影响。分析表明，隔墙提供的阻尼与振幅有关。随着振幅的增大，对阻尼的贡献达到一个平台，超过某一阈值后，阻尼的贡献开始减小。发现隔板增加了结构的固有阻尼比，最大可达8-67 %。在所有情况下，与服务级地震情景相对应的振幅的阻尼比高于考虑的最大地震级别。本研究的结果为地震期间阻尼比的行为提供了有价值的见解，这可以为建筑抗震设计中指定阻尼比的开发指南提供信息。

{"title":"Influence of gypsum partitions on the damping ratio of steel moment frame buildings subjected to earthquakes","authors":"Rosario Jouvhomme , Cristian Cruz","doi":"10.1016/j.engstruct.2026.122114","DOIUrl":"10.1016/j.engstruct.2026.122114","url":null,"abstract":"<div><div>This paper quantifies the variation in modal damping ratios due to the deterioration of gypsum partitions and analyzes their dependence on the amplitude of structural response during earthquakes. A 9-story building with steel moment-resisting frames (SMRF) is modeled in OpenSees, explicitly incorporating gypsum partitions. The steel structure is modeled using linear-elastic elements, while partitions are modeled to account for nonlinear hysteretic behavior. Consequently, the only additional contribution to inherent damping comes from energy dissipation due to partition degradation. A multiple-stripe incremental analysis is conducted to obtain the structural response at different seismic intensities. For each computed response, system identification is employed to determine the first-mode damping ratio and fundamental period of an equivalent linear 9-story structure. The variation in the identified equivalent period and damping ratio with the amplitude of the structural response is then analyzed. A parametric study is conducted to assess the sensitivity of the identified results to: (1) the parameters defining the force-deformation relationship of the partitions; (2) the partition density per floor; (3) the type of partition employed; and (4) the amount of inherent damping in the structure. Finally, three additional SMRF buildings, consisting of 3, 20, and 50 stories are studied to evaluate the influence of building height on the inferred damping ratios. The analysis reveals that the damping provided by the partition walls is amplitude dependent. As amplitude increases, the contribution to damping reaches a plateau, and beyond a certain threshold, it begins to decrease. The partitions were found to increase the inherent damping ratio of the structures by a maximum of 8–67 %. In all cases, the damping ratios at amplitudes corresponding to the service-level earthquake scenario were higher than those at the maximum considered earthquake level. The results of this study offer valuable insights into the behavior of damping ratios during earthquakes, which can inform the development guidelines for specifying damping ratios in the seismic design of buildings.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122114"},"PeriodicalIF":6.4,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940890","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

Study on progressive collapse resistance of steel beam-column structures with column tree connection under edge column failure 边柱破坏下柱树连接钢梁柱结构的抗渐溃性研究

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-06 DOI: 10.1016/j.engstruct.2025.122055

Jian-Gui Qin , Sheng-Yu Lu , Wei-Wei Zhang , Feng Fu , Kai Qian

To investigate the progressive collapse resistance of steel frames with tree connections subjected to edge column failure along the major axis, ensuring the safety of structures under extreme conditions such as impact loads and fires, three full-scale beam-column substructures were designed and tested. The test variables included the presence of lateral restraints and temperature levels. Results indicate that edge column along the major axis failure transforms the substructure into an ultra-long cantilever beam, relying solely on flexural mechanisms for collapse resistance, unlike structures subjected to center column failure which benefit from catenary action. At room temperature, specimens without lateral restraints experienced overall instability, whereas lateral restraints shifted the failure mode to local buckling at the beam end, increasing bearing capacity by 13.1 %. Under high temperatures, laterally restrained substructures still exhibited overall instability due to material degradation. Finite element analysis further revealed that while vertical load magnitude minimally affects failure modes at high temperatures, fire resistance decreases with the vertical load increase on the failed column. When the beam end load increased from 18.68 kN to 32.69 kN, the critical temperature of the substructure decreased by 70℃. Finally, calculation methods for room temperature bearing capacity and load-dependent critical temperature are proposed. Validation against test and FEA results confirmed high accuracy.

为了研究树形连接钢框架在沿主轴边缘柱破坏时的抗倒塌能力，确保结构在冲击载荷和火灾等极端条件下的安全，设计并测试了三个全尺寸梁柱子结构。测试变量包括横向约束和温度水平的存在。结果表明，沿主轴破坏的边缘柱将子结构转变为超长悬臂梁，仅依靠弯曲机制来抵抗倒塌，而中心柱破坏的结构则受益于悬链线作用。在室温下，没有侧向约束的试件整体失稳，而侧向约束将破坏模式转变为梁端局部屈曲，使承载力提高13. %。在高温下，由于材料的降解，侧向约束的子结构仍然表现出整体的不稳定性。有限元分析进一步表明，竖向荷载对高温下破坏模式的影响最小，但随着竖向荷载的增加，破坏柱的耐火性降低。当梁端荷载由18.68 kN增加到32.69 kN时，下部结构临界温度降低70℃。最后，提出了常温承载力和载荷相关临界温度的计算方法。对试验和有限元结果进行了验证。

{"title":"Study on progressive collapse resistance of steel beam-column structures with column tree connection under edge column failure","authors":"Jian-Gui Qin , Sheng-Yu Lu , Wei-Wei Zhang , Feng Fu , Kai Qian","doi":"10.1016/j.engstruct.2025.122055","DOIUrl":"10.1016/j.engstruct.2025.122055","url":null,"abstract":"<div><div>To investigate the progressive collapse resistance of steel frames with tree connections subjected to edge column failure along the major axis, ensuring the safety of structures under extreme conditions such as impact loads and fires, three full-scale beam-column substructures were designed and tested. The test variables included the presence of lateral restraints and temperature levels. Results indicate that edge column along the major axis failure transforms the substructure into an ultra-long cantilever beam, relying solely on flexural mechanisms for collapse resistance, unlike structures subjected to center column failure which benefit from catenary action. At room temperature, specimens without lateral restraints experienced overall instability, whereas lateral restraints shifted the failure mode to local buckling at the beam end, increasing bearing capacity by 13.1 %. Under high temperatures, laterally restrained substructures still exhibited overall instability due to material degradation. Finite element analysis further revealed that while vertical load magnitude minimally affects failure modes at high temperatures, fire resistance decreases with the vertical load increase on the failed column. When the beam end load increased from 18.68 kN to 32.69 kN, the critical temperature of the substructure decreased by 70℃. Finally, calculation methods for room temperature bearing capacity and load-dependent critical temperature are proposed. Validation against test and FEA results confirmed high accuracy.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122055"},"PeriodicalIF":6.4,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940886","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

An efficient single-layer method for estimating the local reliability sensitivity function based on dimension-reduction strategy and sparse grid integration 一种基于降维策略和稀疏网格集成的高效单层局部可靠性灵敏度函数估计方法

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-06 DOI: 10.1016/j.engstruct.2026.122091

Zhuangbo Chen , Zhenzhou Lu

By quantifying the impact of design parameters on failure probability, local reliability sensitivity (LRS) can identify critical parameters and guide reliability based design optimization. Since input variable distribution parameters vary within their design domain, the local reliability sensitivity function (LRSF) can capture this global impact, it is necessary to solve the LRSF. The direct estimation of LRSF requires a double-nested framework of distribution parameter discretization and conditional LRS estimation, resulting in low efficiency. To address this issue, a single-layer method is proposed by combining dimension-reduction strategy and sparse grid integration. The main innovation includes two aspects. The first is constructing an equivalent dimension-reduction formulation to improve the behavior of integrand of the LRSF, enabling the sparse grid integration to efficiently estimate LRSF. The second is introducing a unified density function to formulate a single-layer framework of estimating the LRSF, completely eliminating the double-nested framework of estimating the LRSF. Then, the LRSF can be efficiently estimated by combing the sparse grid strategy based on the Kronrod-Patterson integration rule, high-precision integration nodes and the proposed single-layer solution framework. Both numerical and engineering examples demonstrate the superiority of the proposed method in terms of efficiency, accuracy and robustness.

局部可靠性灵敏度（LRS）通过量化设计参数对失效概率的影响，可以识别关键参数，指导基于可靠性的设计优化。由于输入变量分布参数在其设计域内变化，局部可靠性灵敏度函数（LRSF）可以反映这种全局影响，因此求解LRSF是必要的。直接估计LRSF需要采用分布参数离散化和条件LRS估计的双嵌套框架，效率较低。针对这一问题，提出了一种将降维策略与稀疏网格集成相结合的单层方法。主要创新包括两个方面。首先，构建等效降维公式，改善LRSF的被积函数行为，使稀疏网格积分能够有效地估计LRSF。二是引入统一的密度函数，形成LRSF估计的单层框架，彻底消除了LRSF估计的双嵌套框架。然后，结合基于Kronrod-Patterson积分规则的稀疏网格策略、高精度积分节点和提出的单层解框架，可以有效地估计LRSF。数值和工程实例均证明了该方法在效率、精度和鲁棒性方面的优越性。

{"title":"An efficient single-layer method for estimating the local reliability sensitivity function based on dimension-reduction strategy and sparse grid integration","authors":"Zhuangbo Chen , Zhenzhou Lu","doi":"10.1016/j.engstruct.2026.122091","DOIUrl":"10.1016/j.engstruct.2026.122091","url":null,"abstract":"<div><div>By quantifying the impact of design parameters on failure probability, local reliability sensitivity (LRS) can identify critical parameters and guide reliability based design optimization. Since input variable distribution parameters vary within their design domain, the local reliability sensitivity function (LRSF) can capture this global impact, it is necessary to solve the LRSF. The direct estimation of LRSF requires a double-nested framework of distribution parameter discretization and conditional LRS estimation, resulting in low efficiency. To address this issue, a single-layer method is proposed by combining dimension-reduction strategy and sparse grid integration. The main innovation includes two aspects. The first is constructing an equivalent dimension-reduction formulation to improve the behavior of integrand of the LRSF, enabling the sparse grid integration to efficiently estimate LRSF. The second is introducing a unified density function to formulate a single-layer framework of estimating the LRSF, completely eliminating the double-nested framework of estimating the LRSF. Then, the LRSF can be efficiently estimated by combing the sparse grid strategy based on the Kronrod-Patterson integration rule, high-precision integration nodes and the proposed single-layer solution framework. Both numerical and engineering examples demonstrate the superiority of the proposed method in terms of efficiency, accuracy and robustness.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122091"},"PeriodicalIF":6.4,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145898127","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

UAV-based bridge inspection: RAS-YOLO11 for efficient and accurate multi-class surface defect detection 基于无人机的桥梁检测：RAS-YOLO11，高效准确的多类表面缺陷检测

IF 6.4 1区工程技术 Q1 ENGINEERING, CIVIL

Engineering Structures

Pub Date : 2026-01-05 DOI: 10.1016/j.engstruct.2026.122100

Qiao Chen , Lin Zhou , Zhaolong Bai , Qihang Zhu , Cheng Wang

As key transportation infrastructure, concrete bridges are prone to various types of defects during service, necessitating the use of AI models for efficient detection. Existing object detection methods for surface defect detection in concrete bridges still have key limitations. Some methods select schemes through simple parameter tuning without conducting structural innovations on the model to address practical scenarios, resulting in limited accuracy and complexity. Other models, while incorporating modules such as attention mechanisms to improve accuracy, increase model complexity, making edge deployment difficult, and fail to adapt to devices such as unmanned aerial vehicles (UAVs). To address this, this study constructs a multi-class defect dataset and adopts data augmentation. Based on YOLO11, it integrates innovative feature extraction and downsampling modules into the backbone network, replaces the original neck network with a new one, and proposes a novel model. Experiments demonstrate that the model significantly reduces complexity while improving detection accuracy. After pruning and quantifying the model, it is successfully deployed on resource-constrained edge devices, enabling real-time and efficient recognition of various types of bridge defects. This study achieves dual optimization of accuracy and complexity, overcomes the bottleneck that existing high-precision models are difficult to deploy on the edge, and provides technical support for the application of devices such as UAVs in bridge maintenance.

混凝土桥梁作为关键的交通基础设施，在使用过程中容易出现各种类型的缺陷，需要使用人工智能模型进行高效检测。现有的用于混凝土桥梁表面缺陷检测的目标检测方法仍然存在很大的局限性。有些方法通过简单的参数整定来选择方案，没有针对实际场景对模型进行结构创新，导致精度和复杂性有限。其他模型虽然合并了注意力机制等模块以提高准确性，但增加了模型复杂性，使边缘部署变得困难，并且无法适应无人机（uav）等设备。为了解决这一问题，本研究构建了一个多类缺陷数据集，并采用了数据扩充的方法。在YOLO11的基础上，将创新的特征提取和下采样模块集成到骨干网中，用新的颈部网络代替原有的颈部网络，提出了一种新的模型。实验表明，该模型在提高检测精度的同时显著降低了复杂度。在对模型进行裁剪和量化后，成功地将其部署在资源受限的边缘设备上，实现了对各类桥梁缺陷的实时高效识别。本研究实现了精度和复杂度的双重优化，克服了现有高精度模型难以在边缘部署的瓶颈，为无人机等设备在桥梁维修中的应用提供了技术支持。

{"title":"UAV-based bridge inspection: RAS-YOLO11 for efficient and accurate multi-class surface defect detection","authors":"Qiao Chen , Lin Zhou , Zhaolong Bai , Qihang Zhu , Cheng Wang","doi":"10.1016/j.engstruct.2026.122100","DOIUrl":"10.1016/j.engstruct.2026.122100","url":null,"abstract":"<div><div>As key transportation infrastructure, concrete bridges are prone to various types of defects during service, necessitating the use of AI models for efficient detection. Existing object detection methods for surface defect detection in concrete bridges still have key limitations. Some methods select schemes through simple parameter tuning without conducting structural innovations on the model to address practical scenarios, resulting in limited accuracy and complexity. Other models, while incorporating modules such as attention mechanisms to improve accuracy, increase model complexity, making edge deployment difficult, and fail to adapt to devices such as unmanned aerial vehicles (UAVs). To address this, this study constructs a multi-class defect dataset and adopts data augmentation. Based on YOLO11, it integrates innovative feature extraction and downsampling modules into the backbone network, replaces the original neck network with a new one, and proposes a novel model. Experiments demonstrate that the model significantly reduces complexity while improving detection accuracy. After pruning and quantifying the model, it is successfully deployed on resource-constrained edge devices, enabling real-time and efficient recognition of various types of bridge defects. This study achieves dual optimization of accuracy and complexity, overcomes the bottleneck that existing high-precision models are difficult to deploy on the edge, and provides technical support for the application of devices such as UAVs in bridge maintenance.</div></div>","PeriodicalId":11763,"journal":{"name":"Engineering Structures","volume":"352 ","pages":"Article 122100"},"PeriodicalIF":6.4,"publicationDate":"2026-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145898121","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