Structural Damage Detection Research Articles

Identifying Damage in Structures: Definition of Thresholds to Minimize False Alarms in SHM Systems

In recent years, the development of quick and streamlined methods for the detection and localization of structural damage has been achieved by analysing key dynamic parameters before and after significant events or as a result of aging. Many Structural Health Monitoring (SHM) systems rely on the relationship between occurred damage and variations in eigenfrequencies. While it is acknowledged that damage can affect eigenfrequencies, the reverse is not necessarily true, particularly for minor frequency variations. Thus, reducing false positives is essential for the effectiveness of SHM systems. The aim of this paper is to identify scenarios where observed changes in eigenfrequencies are not caused by structural damage, but rather by non-stationary combinations of input and system response (e.g., wind effects, traffic vibrations), or by stochastic variations in mass, damping, and stiffness (e.g., environmental variations). To achieve this, statistical variations of thresholds were established to separate linear non-stationary behaviour from nonlinear structural behaviour. The Duffing oscillator was employed in this study to perform various nonlinear analyses via Monte Carlo simulations.

Damage detection of road domain waveform guardrail structure based on machine learning multi-module fusion.

The current highway waveform guardrail recognition technology has encountered problems with low segmentation accuracy and strong noise interference. Therefore, an improved U-net semantic segmentation model is proposed to improve the efficiency of road maintenance detection. The model training is guided by mixed expansion convolution and mixed loss function, while the presence of guardrail shedding is investigated by using partial mean values of gray values in ROI region based on segmentation results, while the first-order detail coefficients of wavelet transform are applied to detect guardrail defects and deformation. It has been determined that the Miou and Dice of the improved model are improved by 8.63% and 17.67%, respectively, over the traditional model, and that the method of detecting defects in the data is more accurate than 85%. As a result of efficient detection of highway waveform guardrail, the detection process is shortened and the effectiveness of the detection is improved later on during road maintenance.

Damage detection and location using a simulated annealing-artificial hummingbird algorithm with an improved objective function

Swarm intelligence algorithms and finite element model update technology are important issues in the field of structural damage detection. However, the complexity of engineering structural models normally leads to low computational efficiency and large detection errors in structural damage detection. To solve these problems, a simulated annealing-artificial hummingbird algorithm (SA-AHA) is proposed based on the artificial hummingbird algorithm (AHA). The Sobol sequence is used to improve the identification efficiency by optimizing the initial population distribution of the AHA. Then, the simulated annealing strategy is introduced to improve the detection accuracy by enhancing the global search ability of the AHA. In addition, a novel objective function is presented by combining modal flexibility residual, natural frequency residual, and trace sparse constraint of the structural model. Numerical simulations of a simply supported beam and a two-story rigid frame are carried out to verify the superiority of the proposed SA-AHA and the objective function. Simulation results demonstrate that the SA-AHA is better than the AHA in terms of damage computational efficiency and damage identification accuracy. Moreover, the new objective function can be more excellently applied to the SA-AHA than the previous one, which can be effectively used to locate and estimate the damage of the proposed SA-AHA in structure. Finally, experimental studies are carried out to verify the proposed method.

Distributed fiber optic sensors for structural health monitoring of composite pressure vessels

AbstractIn this paper, we present a comprehensive overview of our research in the field of distributed fiber optic sensors for structural health monitoring of hydrogen composite pressure vessels. Specifically, we demonstrate how the integration of fiber optic sensors into composite pressure vessels enhances safety while simultaneously reducing maintenance costs. The small size of optical fibers enables their integration into composite structures during the manufacturing process, allowing continuous monitoring and precise detection and localization of structural damages during service life. We also discuss the potential of state-of-the-art signal processing methods and machine learning for advancing predictive maintenance. Our applications of fiber optic sensors demonstrate their potential to contribute significantly to the energy transition towards renewable sources.

Assessment of articular cartilage degradation in response to an impact injury using µMRI

ABSTRACT Purpose Degradation of articular cartilage (AC) due to injury to the knee joint may initiate post-traumatic osteoarthritis (PTOA). Failure to diagnose the onset of the disease at an early stage makes the cure ineffective for PTOA. This study investigated the consequences of a mechanical injury to the knee in a rabbit model using microscopic magnetic resonance imaging (µMRI) at high resolution. Materials and methods A mechanical injury was induced to the knee joints of 12 rabbits. Cartilage blocks were extracted from the non-impacted and impacted knee joints after 2 and 14 weeks post-impact. The specimens were studied using µMRI T2 relaxation and inductively coupled plasma analysis to determine the early degradation of the articular cartilage. Results The data established a connection between T2 relaxation time and the early progression of knee PTOA after an impact injury. T2 values were found to be higher in the impacted cartilage at both 2 and 14 weeks, in particular, T2–55° values in the impacted samples displayed a significant rise of 6.93% after 2 weeks and 20.02% after 14 weeks. Lower glycosaminoglycan measurement and higher water content in the impacted cartilage confirmed the µMRI results. Conclusions This µMRI T2 study was able to detect cartilage damage in the impacted knees. In addition, greater degradation in the affected knees at 14 weeks than at 2 weeks indicated the progressive nature of cartilage deterioration over time. The µMRI results were in accord with the biochemical analysis, indicating the detection of early structural damage in the cartilage.

Efficient structural damage detection via the lω regularization and randomized extended Kaczmarz algorithm

Structural damage detection (SDD) is an important aspect of structural health monitoring. This study aimed to explore a new method IT ω -REK θ (ω = 1, 1/2; θ = 1,2,3) for SDD based on the lω sparse regularization model and the randomized extended Kaczmarz (REK) type algorithms. When ω = 1/2, the l1/2 sparse regularization model was applied to enhance the ill-posedness of the damage identification problem and ensure the sparsity of the solution. The REK, θ = 1, partially randomized extended Kaczmarz (θ = 2), and fast maximum-distance extended Kaczmarz (θ = 3) algorithms with different threshold operators were used to solve the lω regularization model. These algorithms could obtain optimal identification results and significantly improve computational efficiency by randomly and partially selecting the data of the sensitivity equations. Numerical and experimental studies on different structures showed that the proposed method could fast locate structural damage and accurately identify the damage extents, which was robust to the SDD problem with noise.

Acoustic emission detection of filament wound CFRP composite structure damage based on Mel spectrogram and deep learning

This paper investigates the hoop tensile mechanical properties, damage mode classification, and damage evolution process of the filament wound CFRP composite structure with different winding angles by acoustic emission and deep learning. Firstly, a clear correspondence between damage modes and peak frequency ranges is established by using k-means clustering results and eight frequency band ranges obtained by wavelet packet transform. On this basis, three damage modes are characterized according to the energy content of different frequency bands. Subsequently, the trained Mel spectrum-CNN classification model is used to identify damage modes and analyze the damage evolution process. The results show that the hoop tensile strength of the filament wound CFRP composite structure with ±55° is higher than that with ±35°. The primary damage modes during the tensile process are matrix cracking and fiber/matrix debonding. The Mel spectrogram-based damage identification method can effectively classify damage modes. In addition, the combination of acoustic emission and deep learning shows significant potential for damage identification and monitoring.

Crack detection in historic structures using an improved Prewitt filter: Case study of the Historic Si-o-se-pol Bridge in Iran

In contrast to new and modern structures, historical structures have cultural significance, and it is essential to make an effort to preserve them. For this purpose, fast and accurate damage detection of historical structures plays a crucial role. Thus, the current study introduces a novel and effective tool for detecting cracks in historic structures. A new and modified version of the Prewitt filter is proposed to detect cracks in the Si-o-se-pol Bridge, a historic bridge in Iran. Results of crack detection using vertical and horizontal Prewitt filters are compared to those obtained from the proposed improved filter. The findings show that the improved Prewitt filter has significantly better performance compared to the vertical and horizontal Prewitt filters.

Detecting damages in metallic beam structures using a novel wavelet selection criterion

Structural damage detection using wavelet transform is effective if a suitable wavelet function is selected. However, the selection of the appropriate wavelet function can be a challenging task. This paper introduces a novel and efficient criterion for selecting the proper wavelet function. It is emphasized that there is a relationship between the accuracy of damage detection using wavelet detail coefficients and the correctness of the approximation of the original function. Accordingly, we define a ratio function and then a novel wavelet selection criterion called WSC in this study. After applying the proposed WSC criterion for 128 wavelet functions, it is found that the wavelet functions selected via the proposed criterion (both in numerical and experimental scenarios) can detect the position of damages in the beam structure with high accuracy.

Damage Detection in Structures by Using Imbalanced Classification Algorithms

Detecting damage constitutes the primary and pivotal stage in monitoring a structure’s health. Early identification of structural issues, coupled with a precise understanding of the structure’s condition, represents a cornerstone in the practices of structural health monitoring (SHM). While many existing methods prove effective when the number of data points in both healthy and damaged states is equal, this article employs algorithms tailored for detecting damage in situations where data are imbalanced. Imbalance, in this context, denotes a significant difference in the number of data points between the healthy and damaged states, essentially introducing an imbalance within the dataset. Four imbalanced classification algorithms are applied to two benchmark structures: the first, a numerical model of a four-story steel building, and the second, a bridge constructed in China. This research thoroughly assesses the performance of these four algorithms for each structure, both individually and collectively.

Few-shot meta transfer learning-based damage detection of composite structures

Damage detection and localization using data-driven approaches in carbon fiber reinforced plastics (CFRP) composite structures is becoming increasingly important. However, the performance of conventional data-driven methods degrades greatly under little amount of data. In addition, the scarcity of data corresponding to defect/damage conditions of CFRP structures lead to extreme data imbalance, which make this problem even more challenging. To address these challenges of few training data and the scarcity of damage samples, this paper proposes a few-shot meta transfer learning (FMTL)-based approach for damage detection in CFRP composite structures. This method leverages knowledge learnt from an unbalanced data domain generated from a single CFRP composite sample and adapts the knowledge to be applied for other data domains generated by CFRP samples with different structural properties. The contributions of this research include demonstrating the feasibility of harnessing knowledge from notably limited experiment data, designing an algorithm for configuring hyperparameters based on a specific FMTL task, and identifying the impacts of hyperparameters on learning performances. Results show that FMTL can improve the recall rate by at least 15% while preserving the ability to identify health conditions. This method can be extremely useful when we need to monitor health condition of critical CFRP structures, like airplanes, because they can rarely generate data under damage conditions for model training. FMTL enables us to build new models based on unbalanced source domain data with the cost of a minimal set of samples from the target domain.

Magnetic resonance neurography in spinal cord injury: Imaging findings and clinical significance.

It is unknown whether changes to the peripheral nervous system following spinal cord injury (SCI) are relevant for functional recovery or the development of neuropathic pain below the level of injury. Magnetic resonance neurography (MRN) at 3 T allows detection and localization of structural and functional nerve damage. This study aimed to combine MRN and clinical assessments in individuals with chronic SCI and nondisabled controls. Twenty participants with chronic SCI and 20 controls matched for gender, age, and body mass index underwent MRN of the L5 dorsal root ganglia (DRG) and the sciatic nerve. DRG volume, sciatic nerve mean cross-sectional area (CSA), fascicular lesion load, and fractional anisotropy (FA), a marker for functional nerve integrity, were calculated. Results were correlated with clinical assessments and nerve conduction studies. Sciatic nerve CSA and lesion load were higher (21.29 ± 5.82 mm2 vs. 14.08 ± 4.62 mm2 , p < 0.001; and 8.70 ± 7.47% vs. 3.60 ± 2.45%, p < 0.001) in individuals with SCI compared to controls, whereas FA was lower (0.55 ± 0.11 vs. 0.63 ± 0.08, p = 0.022). DRG volumes were larger in individuals with SCI who suffered from neuropathic pain compared to those without neuropathic pain (223.7 ± 53.08 mm3 vs. 159.7 ± 55.66 mm3 , p = 0.043). Sciatic MRN parameters correlated with electrophysiological results but did not correlate with the extent of myelopathy or clinical severity of SCI. Individuals with chronic SCI are subject to a decline of structural peripheral nerve integrity that may occur independently from the clinical severity of SCI. Larger volumes of DRG in SCI with neuropathic pain support existing evidence from animal studies on SCI-related neuropathic pain.

Enhancing Seismic Damage Detection and Assessment in Highway Bridge Systems: A Pattern Recognition Approach with Bayesian Optimization.

Highway bridges stand as paramount elements within transportation infrastructure systems. The ability to ensure swift recovery after extreme events, such as earthquakes, is a fundamental trait of resilient communities. Consequently, expediting the recovery process necessitates near real-time diagnosis of structural damage to provide dependable information. In this study, a data-driven approach for damage detection and assessment is investigated, focusing on bridge columns-the pivotal supporting elements of bridge systems-based on simulations derived from nonlinear time history analysis. This research introduces a set of cumulative intensity-based damage features, whose efficacy is demonstrated through unsupervised learning techniques. Leveraging the support vector machine, a prominent pattern recognition algorithm in supervised learning, alongside Bayesian optimization with a Gaussian process, seismic damage detection and assessment are explored. Encouragingly, the methodology yields high estimation accuracies for both binary outcomes (indicating the presence of damage or the occurrence of collapse) and multi-class classifications (indicating the severity of damage). This breakthrough opens avenues for the practical implementation of on-board sensor computing, enabling near real-time damage detection and assessment in bridge structures.

Arbitrary-Order Sensitivity Analysis in Wave Propagation Problems Using Hypercomplex Spectral Finite Element Method

Many modern structural health monitoring (SHM) systems use piezoelectric transducers to induce and measure guided waves propagating in structures for structural damage detection. To increase the detection capabilities of SHM systems, gradient-based optimization of sensor placement is frequently necessary. However, available numerical differentiation methods for mechanical wave propagation problems suffer from truncation and subtraction errors and are difficult to extend to high-order sensitivities. This paper addresses these issues by introducing an approach to obtain highly accurate numerical sensitivities of arbitrary order in mechanical wave propagation problems. The hypercomplex time-domain spectral finite element method (ZSFEM) couples the hypercomplex Taylor series expansion method with the time-domain spectral finite element method. We show how ZSFEM can be implemented within the commercial finite element package ABAQUS/Explicit. For verification, we compared the numerical and analytical results of the displacement and its sensitivities with respect to mechanical parameters, geometry, and boundary conditions for a rod subjected to a sudden, distributed axial load. First- and second-order sensitivities were obtained with normalized root mean square deviations below 4×10−3. Mesh convergence analyses revealed that p-refinement offered better convergence rates than h-refinement for the outputs and their sensitivities. Also, the sensitivities obtained with ZSFEM were compared with finite differences showing higher accuracy and step-size independence (e.g., no iteration is needed to determine the step size that minimizes the error). For simplicity, ZSFEM was presented only for one-dimensional truss elements, but the method is general and can be applied to other elements.

Towards vibration-based damage detection of civil engineering structures: overview, challenges, and future prospects

AbstractIn this paper, we delve into the evolving landscape of vibration-based structural damage detection (SDD) methodologies, emphasizing the pivotal role civil structures play in society's wellbeing and progress. While the significance of monitoring the resilience, durability, and overall health of these structures remains paramount, the methodology employed is continually evolving. Our focus encompasses not just the transformation brought by the advent of artificial intelligence but also the nuanced challenges and future directions that emerge from this integration. We shed light on the inherent nonlinearities civil engineering structures face, the limitations of current validation metrics, and the conundrums introduced by inverse analysis. Highlighting machine learning's (ML) transformative role, we discuss how techniques such as artificial neural networks and support vector machine's have expanded the SDD's scope. Deep learning's (DL) contributions, especially the innovative capabilities of convolutional neural network in raw data feature extraction, are elaborated upon, juxtaposed with the potential pitfalls, like data overfitting. We propose future avenues for the field, such as blending undamaged real-world data with simulated damage scenarios and a tilt towards unsupervised algorithms. By synthesizing these insights, our review offers an updated perspective on the amalgamation of traditional SDD techniques with ML and DL, underlining their potential in fostering more robust civil infrastructures.

Moving-Principal-Component-Analysis-Based Structural Damage Detection for Highway Bridges in Operational Environments.

With the deterioration of bridge performance and ever-increasing amounts of traffic, bridge safety is becoming a concern for the engineering community. A method that can assess a bridge's condition in real time is urgently needed. The main factors that hinder bridge condition assessment are the uncertain operational environments. A new moving principal component analysis (MPCA)-based method is developed for structural damage detection in bridges in operational environments in this paper. Two main operational environmental factors, the environmental temperature and traffic loads, are studied in the assessment process to verify the robustness and practicality of the proposed method. The numerical and experimental results show that the proposed method is effective and accurate in assessing the bridge's condition in operational environments.

Structural damage detection of 3-D truss structure using nodal response analysis

Structural damage detection of 3-D truss structure using nodal response analysis

Physics-guided deep learning for damage detection in CFRP composite structures

Structural health monitoring of carbon fiber reinforced plastics composite structures based on deep learning usually builds on black box models for damage detection. However, the lack of physics knowledge limits its application into real-world problems where structural property or environmental condition varies. To overcome this limitation, we propose a physics-guided deep learning framework to integrate physics into data-driven models. This physics-guided convolutional neural network leverages structural degradation trend and physical consistency by combining the output of the physical model with the observed feature in a hybrid model. This hybrid model uses an additional branch to observe the information of stiffness degradation, which is the input into the physical model to describe the damage growth in structures by establishing a relationship with the power spectral density change in the guided wave signals. Additionally, the physics-based loss function is designed as part of the learning objective to ensure the model outputs satisfy the existing physics and maintain physical consistency. Experiment results show that our method demonstrates great generalizability that by training on data of only one particular CFRP composite structure, the PGCNN model can expand its great performance on all other CFRP layups in the NASA-published CFRP Dataset with over 90% accuracy.

A Substructure-based Model Updating Method to Damage Detection for Large-scale Structures

The health monitoring and damage detection of structures are major scientific issues that the civil engineering discipline, which made great achievements in the 20th century, has bequeathed to the 21st century. Despite the installation of health monitoring systems in many large structures, implementing damage detection through finite element model updating is often time-consuming or even infeasible due to the size of the model and the presence of numerous uncertain parameters. To address this issue, this paper proposes a substructure-based model updating method. The entire structure is divided into substructures with reduced degrees of freedom, allowing for the simplification of the motion equation by employing only a small number of low-order modes from each substructure. Consequently, the analysis scale of the structure is effectively reduced. The discarded higher modes are compensated by residual modes to ensure the accuracy of structural response and sensitivity. Subsequently, a damage identification program substructure-based model updating is developed, which is applied to component mode synthesis and damage identification for large structures. By precisely detecting damage in critical areas, accurate diagnosis and evaluation of the global structural safety are achieved. The results validate the implementation, computational efficiency, and accuracy of the proposed substructure-based model updating method. This approach shifts the focus from model updating of the entire structure to model updating of substructures, with the aim of fundamentally resolving the technical challenge of accurate damage detection of large and complex structures. Furthermore, it provides theoretical support for the practical application of damage detection in large civil structures.

Damage detection of FRP structures using the electrical resistance response of MWCNTs networks in a PDMS film substrate

Damage detection of FRP structures using the electrical resistance response of MWCNTs networks in a PDMS film substrate