Structural Control & Health Monitoring最新文献

As a pivotal component in advancing the sustainable development goals (SDGs), structural health monitoring (SHM) has garnered increasing attention in the field of civil engineering. Considering the various approaches, model-based SHM is the most prevalent and remains highly effective due to its theoretical framework and nondestructive nature, creating a robust framework for effective SHM, enabling early detection of issues, and supporting informed maintenance strategies. Through decades, stochastic subspace identification (SSI) has been proven, and recursive SSI (RSSI) has been consequently applied for model-based SHM due to its ability to track modal parameters and generate accurate models. However, online validation through structural experiments has yet to be conducted with large-scale specimens. In this study, a shaking table experiment is conducted to validate the online implementation of RSSI for tracking time-varying modal parameters in real time. The full-scale specimen, experimental setup, and test framework are first described with great detail, and a numerical model is developed through a pretest using the shaking table system located in Taiwan. Subsequently, the simulation study provides numerous suggestions for experimental implementation. The experimental study then demonstrates that the proposed approach not only enables an online identification but also produces an accurate dynamic model. Besides, practical measures are recommended to fulfill online processing through the comprehensive simulation and experiential studies, especially those related to the user-defined parameters and ambient excitations. The results evidence that the SHM systems based on RSSI can effectively track the changes of dynamic characteristics under ambient excitations, ultimately facilitating the assessment and maintenance of structures.

Detecting and classifying road damage are crucial for road maintenance. To address the limitations of existing road damage detection methods, including insufficient fine-grained contextual feature extraction and complex models unsuitable for deployment, this paper proposes a lightweight backbone and neck road damage detection model named LBN-YOLO. First, the backbone and neck of the original model are improved to be lightweight, and the C2f-dilation wise residual (C2f-DWR) module is integrated in the backbone to extract multiscale contextual information. Second, a simplified bidirectional feature pyramid network is employed in the neck structure to optimize the feature fusion network, reducing the number of parameters and simplifying the model complexity. Finally, a dynamic head with self-attention is introduced to enhance the sensing capability of the detection head, thus improving the precision of detecting occluded small objects. The proposed model’s detection ability is evaluated using a custom road damage dataset. The experimental results demonstrate that our proposed LBN-YOLO model achieves superior performance compared with the YOLOv8n model, with an increase of 4.1% in [email protected] and a 5.2% enhancement in precision, outperforming other detection models. In addition, the model is evaluated on two public datasets, showing improved detection performance compared with the original model, demonstrating strong generalization capabilities. Code and dataset are available at https://github.com/gzNiuadc/Road-crack-dataset.

The preservation of historical bridges usually requires extensive structural evaluations for possible damage detection. Therefore, effective techniques are essential for diagnosis and, consequently, proper maintenance and rehabilitation actions. The combination of techniques provides complementary data that support decision making. A complete assessment was applied in the study of the Comboa Bridge, a medieval masonry structure in river Verdugo, in Galicia (Spain). It has three irregular arches, and the first visual inspection denotes the existence of important cracking and vegetation in the stonework. One of the most representative nondestructive testing (NDT) techniques for in situ evaluation is ground-penetrating radar (GPR) that offers detailed insights into subsurface conditions, revealing information about materials, voids, and deterioration, while loading tests and passive seismic methods provide dynamic responses that are related to type of structure and possible damage. This method was combined with loading tests to obtain deflections of the bridge deck and passive seismic for analyzing the dynamic behavior. Moreover, 3D models of the structure were set up using light detection and ranging (LiDAR), performed with terrestrial laser scanning, and unmanned aerial vehicle (UAV) surveying. By combining 3D models with NDT techniques, the results provide comprehensive information that enhances the understanding of a bridge’s condition and safety. These results are used for calibrating the dynamic computational model of the structure in order to obtain the vibration modes. Each technique used in the study presents limitations, which are addressed and discussed herein. Furthermore, the site conditions can also affect the results, as the effectiveness of these methods can vary greatly, depending on the materials and structures, which influences the electromagnetic and mechanical wave propagation. Additionally, the frequency of the waves may not effectively mark all relevant structural features or smaller damage. When used together, the NDT methods can complement each other’s strengths, but challenges remain. Overall, while these techniques are valuable tools for assessing historical bridges, awareness of their limitations is crucial for accurate interpretation and effective decision making in preservation efforts. The results obtained in the Comboa Bridge demonstrate improved accuracy in identifying structural anomalies. Additionally, recommendations to overcome some of these challenges in case of historical bridge assessment and also for the continuous monitoring and adequate maintenance actions to preserve the bridge integrity and safety are presented.

As subway train-induced low-frequency vibrations continue to rise, there is an increasing need for more effective vibration control strategies. Although current low-frequency vibration reduction methods offer some solutions, further progress is necessary. This paper introduces a novel tuned liquid particle damper-dynamic vibration absorber (TLPD-DVA), which merges the principles of tuned liquid dampers (TLDs) and particle dampers (PDs). By capitalizing on the low-frequency damping capabilities of TLDs, this approach incorporates particles suspended within the liquid to create a hybrid damping device capable of effectively attenuating vibrations across a wide low-frequency range (10 to 80 Hz). A discrete element method-computational fluid dynamics (DEM-CFD) model for multiphase flow is employed to explore the damping mechanism, optimize system parameters, and develop a frequency-dependent nonlinear damping device. The TLPD-DVA is then applied to floating slab track systems to control low-frequency vibrations, and a dynamic interaction model involving the coupled vehicle-TLPD-DVA-floating slab track-tunnel system is established to assess the system’s response. Harmonic response analysis of a floating slab track fitted with TLPD-DVAs, along with dynamic mass and mass ratio indices, clarifies the vibration reduction mechanism. Additionally, field tests demonstrate that the TLPD-DVA reduces vertical acceleration on the floating slab by up to 8 dB and on the tunnel wall by up to 10 dB within the low-frequency range, surpassing the performance of tuned DVAs. The proposed TLPD-DVA offers significant potential for vibration control in various civil engineering applications, including transportation infrastructure, building foundations, and vibration-sensitive facilities.

The bridge structural health monitoring (SHM) system will inevitably experience missing data. To ensure the integrity and practicability of the bridge SHM system, it is essential to repair the missing data. The existing data recovery methods mainly use the spatial correlation with other monitoring data but cannot adequately capture the time dependence of the raw monitoring data. This paper uses historical monitoring data to predict future data and complete the task of repairing missing data. A hybrid prediction model based on the gated recurrent unit (GRU) neural network, complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), and variational mode decomposition (VMD) is proposed. By decomposing the raw monitoring data, the input of the GRU model is optimized, resulting in improved accuracy of prediction and enabling the model to operate independently from other sensors. The accuracy of the method is verified based on the SHM data of a cable-stayed bridge. The prediction results of the proposed model are stable and reliable, with a prediction accuracy reaching 95%, indicating that the CEEMDAN-VMD-GRU model is suitable for repairing missing deflection data in bridge SHM systems.

Self-heating effect of the lead core in lead–rubber bearings (LRBs) under cyclic loading causes degradation of mechanical properties of LRBs, which in turn affects their self-heating effect. This study conducts full-scale tests and proposes a numerical modeling methodology to investigate the coupled thermal and mechanical behavior of LRBs. The methodology integrates mechanical modeling, thermal modeling, temperature-dependent material properties, and thermal-mechanical modeling. Experimental results reveal significant mechanical degradation under high-speed cyclic loading (0.25 Hz, 100% shear strain), with a temperature rise of 90°C in the lead core and a 22°C increase observed in adjacent rubber layers after 10 cycles. The numerical model demonstrates a good agreement with test data, accurately capturing force-displacement loops and temperature within the lead core. Numerical results show that the thermal–mechanical behavior of LRBs is sensitive to loading frequency and shear strain: increasing the frequency from 0.25 Hz to 0.5 Hz amplifies energy dissipation rates by 38%, while a 50% increase in shear strain (100%–150%) increases peak temperatures by 27%. A case study under nonharmonic motion shows that conventional mechanical models overestimate energy dissipation by 37% compared to the coupled thermal–mechanical model. The proposed modeling methodology provides a usable tool for investigating the coupled thermal and mechanical behavior of LRBs under various seismic conditions.

Dual-column bents with energy dissipation components represent one structural type of seismic resilient bridge bents. However, there have been few dynamic analyses and shaking table tests to validate the theoretical formulas, design methodology, and the seismic effects of the energy dissipation elements on dual-column bents, particularly for the tall dual-column bents. Thus, the study aims to derive theoretical formulas for calculating the yield strength, yield displacement, and elastic stiffness of tall dual-column bents with and without shear beams (SBs). This enables a more comprehensive understanding of structural performance and allows for more accurate predictions of the seismic behavior of these novel bents during seismic events compared to existing studies. A structural fuse-based design methodology for tall dual-column bents was developed and validated through verified finite element models and shaking table tests. Both numerical and experimental analyses were conducted to evaluate the effectiveness of SBs in mitigating seismic damage and responses in tall dual-column bents. The displacements and curvatures of the tall dual-column bents with and without SBs were analyzed. The results show that SBs enhance the seismic resilience and decrease the seismic responses of tall dual-column bents by strategically yielding first to dissipate energy, achieving up to 69.5% displacement reduction under E1-level earthquakes (PGA = 0.40 g) and 77.6% curvature reduction under E2-level earthquakes (PGA = 0.68 g) compared to tall dual-column bents without SBs. Crucially, this prioritized yielding mechanism enables SBs to function as structural fuses, suppressing structural responses below critical yield thresholds and safeguarding columns. Consequently, the tall dual-column bent without SBs undergoes seismic damage under E1-level earthquakes, while the tall dual-column bent with SBs does not suffer any damage. SBs enable the tall dual-column bent to meet performance targets. This suggests that SBs notably enhance the seismic resilience of tall dual-column bents, and the proposed design method can be used to design actual engineering structures.

Zero-shot learning approaches have emerged as promising techniques for structural health monitoring (SHM) due to their ability to learn representations without labeled data. With the practical design of such models, the shift from traditional structure-dependent techniques to potentially large-scale implementations becomes feasible, effectively addressing the challenge of gathering labeled data. Autoencoders (AEs), a class of deep neural networks, align well with zero-shot SHM settings due to their architecture, loss function, and optimization process. In AEs, the reconstruction error is expected to increase for novel data patterns (i.e., potential damage data), while the encoded manifold in their bottleneck layers enables the discrimination of complex patterns. However, for practical SHM applications, rigorous evaluation of (variational) AEs and the robustness of reconstruction loss- or manifold-based designs in handling real-world scenarios remains necessary. Accordingly, this article employs two SHM benchmarks to evaluate the effectiveness of manifold learning compared to the reconstruction errors of (variational) AEs in a zero-shot setting. The comparison encompasses metrics such as reconstruction fidelity, preservation of structural characteristics, and the ability to generalize to unseen structural conditions. Furthermore, an unbiased normalization-based ensemble methodology is proposed, combining both approaches with the goal of enhancing damage detection performance and delivering more reliable results in zero-shot learning contexts. The proposed ensemble strategy, integrating both reconstruction error and manifold representations, adds robustness to the damage detection process, a crucial feature in the uncertain domain of zero-shot structural damage detection. The findings suggest that neither reconstruction loss nor manifold data consistently outperform the other; structural differences may render one approach more effective than the other in specific contexts, and based on these observations, a zero-shot damage severity index is suggested and tested on the benchmark data. Nevertheless, the proposed ensemble method demonstrates superior performance over individual models in estimating damage severity in an unsupervised setting. These results highlight the efficacy of variational AEs for zero-shot SHM, offering insights into their strengths and limitations and aiding users in selecting appropriate zero-shot damage detection strategies in the absence of labeled data.

Accurate prediction of bridge structural responses is crucial for infrastructure safety and maintenance. This study introduces the Nonstationary Transformer (NSFormer), a novel model designed to address the challenges posed by nonstationary data in bridge monitoring, characterized by trends, periodicity, and random fluctuations. Unlike traditional models such as LSTM and Transformer, NSFormer leverages a de-stationary attention mechanism that dynamically adapts to changing temporal patterns, enabling robust long-term prediction. Experimental results show that NSFormer consistently outperforms the traditional models across multiple datasets and prediction horizons. Specifically, at a 24-step prediction horizon, NSFormer reduces mean absolute error by at least 22.88% for Deflection dataset and 66.67% for Strain-All dataset. While predictive accuracy decreases with longer horizons, NSFormer maintains superior performance compared to alternatives. Furthermore, prediction accuracy remains stable across varying input horizons, demonstrating the model’s ability to effectively capture temporal dependencies despite data variability. These findings imply that NSFormer can significantly enhance the reliability of structural health monitoring systems by providing more accurate and stable prediction under complex, variable conditions, thereby supporting timely maintenance decisions and improving bridge safety management.

Damping ratio estimation for bridges under operational conditions typically employs operational modal analysis (OMA) methods. However, existing comparisons of these methods often overlook the nonstationary nature of traffic loads. This study focuses on two key aspects: (1) the performance evaluation of four OMA methods, autocorrelation function (ACF), stochastic subspace identification (SSI), random decrement technique (RDT), and decay response extraction (DRE), under nonstationary traffic loading, and (2) the quantification of the effects of temperature, traffic load, and wind load on structural damping ratios. An automatic modal parameter identification approach was developed to analyze two-year monitoring data from a single-tower cable-stayed bridge. The practical performance of each method was assessed statistically. Finally, a method was proposed to separate the effects of temperature and traffic loading at different time scales, and a damping ratio prediction model was established. The results indicate that both SSI and ACF methods demonstrate good performance, with the ACF method exhibiting smaller variance. SSI requires careful handling of false modes, RDT has the largest variance, and the DRE method suffers from uneven temporal distribution of identification results. Temperature and traffic loading have significant effects on the damping ratios of the bridge.