Structural Damage Detection Research Articles

A comparison of modal parameter identification tests in laboratory conditions

AbstractModal parameter evaluation is widely used for structural health monitoring, damage detection, and numerical model calibration. This study benchmarks two methods for modal identification under controlled laboratory conditions using operational modal analysis to compare dynamic parameters from ambient vibrations and shaker tests. A 2‐storey 3D spatial steel frame equipped with 3‐axial accelerometers is used as a case study. The identified dynamic parameters are benchmarked against a preliminary finite element model by computing Auto‐MAC and MAC matrices. It can be concluded that the preliminary FEM approach performs better in the prediction of modes shapes than natural frequencies. It was found that analysing the structural response to both ambient and artificially induced excitations provide a more complete understanding of a structure's dynamic characteristics. An additional mode shape was identified from the artificially induced excitation that was not previously captured from the ambient excitation.

Inverse Analysis and Optimization‐based Model Updating for Structural Damage Detection

AbstractStructural health monitoring and early detection of structural damage is extremely important to maintain and preserve the service life of civil engineering structures. Identification of structural damage is usually performed using non‐destructive vibration experiments combined with a mathematical procedure called model updating. The finite element model of the investigated structure is updated by incrementally adjusting its parameters so that the model responses gradually approach those of the real possibly damaged structure under investigation. This paper describes the use of two model updating methods. The first method employs metaheuristic optimization technique aimed multilevel sampling to efficiently search the design parameter space to achieve the best match between the deformed structure and its model. The second method approaches model updating as an inverse problem and uses machine learning‐based model to approximate inverse relationship between structural response and structural parameters. Both methods are applied to damage identification of single‐ and double‐span steel trusses. Finally, initial results of the hybrid method are presented. The effect of the damage rate and location on the identification speed and the accuracy of the solution is investigated and discussed.

A wavelet based data coupling method for spatial damage detection in beam-type structures.

Spatial damage identification is of great significance in mechanical, aerospace, and civil engineering. In this study, a data coupling method based on continuous wavelet transform (CWT) is proposed to identify the spatial damage location of beam-type structures. The singularity of the wavelet coefficient can be used to identify the signal singularity, and data coupling method calculates the spatial location of the damage. Numerical simulations and experimental analyses of different type of beams with transfixion damage are carried out to evaluate the accuracy of the method. The results show that the wavelet based data coupling method (W-DCM) can identify the minimum 4.9% damage severity of fixed beam and continuous beam, and can also identify the damage of non-free end of cantilever beam. However, the 9.7% damage severity of the free end of the cantilever beam cannot be identified. It is also found that the W-DCM can effectively circumvent the problem of wavelet coefficients edge effect. This method and wavelet singularity are used to provide a solution to the problem of structural edge damage identification.

A Novel Approach for Structural Damage Detection Using Multi-Headed Stacked Deep Ensemble Learning

A Novel Approach for Structural Damage Detection Using Multi-Headed Stacked Deep Ensemble Learning

Retracted: Application of Particle Swarm Algorithm in Nanoscale Damage Detection and Identification of Steel Structure.

[This retracts the article DOI: 10.1155/2022/4300840.].

Damage detection of structures based on wavelet analysis using improved AlexNet

Deep learning-based approaches have garnered a great deal of interest among different methods in structural health monitoring (SHM), whose primary objective is to assess structural health status. AlexNet is renowned for being a practical convolutional neural network architecture utilized to classify images. However, having numerous parameters, AlexNet gives rise to significant computational costs when applied to classify a small number of damage types occurred in civil engineering. Therefore, in this paper, the improved AlexNet with a lower computational cost and images generated by continuous wavelet transform (CWT) have been employed to classify multiple types of damages which is the pivotal characteristics of the proposed procedure. The performance of this approach was validated with responses obtained from three structures. The acceleration responses converted to time–frequency images were derived from undamaged and damaged states, such as scour, damage in connections, and stiffness reduction. Moreover, the efficiency in detecting the damage extent was investigated. The approach exhibited proficient performance in multi-classification, achieving high prediction accuracy rates of more than 99%, 97%, and 94% in a benchmark bridge, a frame model, and a real truss bridge, respectively. The sensitivity analysis results showed that the Bump mother wavelet had the best performance among studied mother wavelets; moreover, investigation of the number of input images showed that the higher the number of pictures, the more accurate the results, and prediction accuracy fluctuated by no more than approximately 5%. Finally, results demonstrated that the accuracy improves as the duration of each record increases, and the related decrease in accuracy is roughly 3%.

Damage detection in ship hull structures under operational variability through strain sensing

Recently, growing concern for the reliability of ship structures has emerged among stakeholders in the maritime sector. In the transition towards predictive maintenance, Structural Health Monitoring (SHM) has emerged as a viable option. SHM systems seek to infer the existence or level of structural degradation using response data from in-situ sensors. In this work, damage detection in marine stiffened panels is treated. Out-of-plane deflections of the panel’s plate elements act as the damage case of interest, due to being potential triggers for buckling collapse. Strain response data are employed as the damage-sensitive features. Two different approaches are employed for damage detection. In the first, the problem is cast as one of multi-class classification, where the different classes correspond to discrete deflection levels, and a detection theory-based classifier is employed. In the second, the problem is treated as one of probabilistic regression, where strain readings are mapped to the probability distribution of the deflection magnitude. Both methods are implemented on a realistic stiffened panel geometry subjected to operational variability, based on an existing vessel, while all the employed data is obtained through Finite Element (FE) simulations. Results indicate that the proposed architecture can effectively provide uncertainty-informed predictions of out-of-plane deflection levels.

Combination of Cross-Correlation-Based Analysis and Ultrasonic GW Tomography for Barely Visible Impact Damage Detection Preliminary Assessment

Statistical based reconstruction methods and signal processing tooling techniques are implemented and used to detect delaminations or debondings within composite complex items with very high precision. From the literature, it appears that although a single procedure for the estimation of structural health is a fast solution, multiple analyses based on different reconstruction methods or different damage parameters are the way to achieve maturation assessments of the methodology. This highlights the fact that the hardware and software parts of an SHM system need two different assessment and maturation ways. This work focuses on the software part by proposing a way to start assessing the outcomes. In this paper, the damage detection and localization strategy in CFRP plate-like structures with elastic guided waves excited and acquired with a circular array PWAS network is considered. Previous outcomes are compared by new analyses using a new post-processing approach based on a cross-correlation-based technique in terms of the BVID (Barely Visible Impact Damage) surface position and its center of mass. The advantage of this specific study is hopefully to enable confidence in the transition from R&D to field implementation. In addition, this work tries to evidence an improvement in terms of cost efficiency and reduced complexity while maintaining the same accuracy.

Applications of Linear Scanning Magnetostrictive Transducers (MST) for Finding Hard-to-Detect Anomalies in Structural Components

Guided wave testing is now a widely accepted method for detection of structural damage in many different types of components, from pipelines to pressure vessels to tanks. Torsional wave modes (T modes) in pipes and shear horizontal (SH) mode guided waves in plates are good candidates for finding areas with generalized corrosion, due to the absence of fluid coupling effects and their lack of dispersion. However, from our field test experience, certain types of defects are difficult to detect with conventional T mode or SH mode guided wave probes. Gradual wall thinning is one such type of defect; another is crack-like defects in or close to welds or penetrations in the pipe. Recently, Southwest Research Institute (SwRI) has developed a new sensor configuration and scanning system that overcomes these limitations. We have recently developed a linear scanning magnetostrictive transducer (MsT) probe system, in which a FeCo strip wound with radio frequency (RF) coils is attached to the structure under test with shear wave couplant, and a moving permanent magnet driven by a motor is used to excite SH guided waves at predefined positions along this strip. The probe is designed to operate over a wide frequency range (20 – 500 kHz) and can be used to generate dispersive shear wave mode (SH1) in addition to the nondispersive SH0 mode. The use of different modes with the sensor at multiple positions allows detection of a range of defect types. In this paper, the performance of linear scanning MsTs is presented, including experimental evaluation of detection of gradual wall thinning patches of different depths and locations on a steel pipe and plate mockups. Another set of experiments evaluates detection of notches in seam welds. Indications from real-time B-scan and SAFT (synthetic aperture focusing technique) processing will be presented.

Baseline-free structural damage detection using PCA- Hilbert transform with limited sensors

Conventional vibration-based damage identification methods need a large number of sensors on the bridge to identify the structural modal information and rely on the response data of bridges under a non-damage state for comparison, which restricts their implementations in the health monitoring of bridge structures in service. This paper proposes a baseline-free structural damage detection method based on limited sensor information. The proposed method uses the Hilbert transform combined with principal component analysis (PCA) to locate the damage. Firstly, PCA is performed on the displacement response data matrix of the bridge under a moving load to obtain the principal component matrix. According to the physical interpretation of PCA of the responses of a beam-like bridge under a moving load, each column of the principal component matrix, namely the principal component, composes of a mode shape and the dynamic information at this modal frequency of the bridge. The dynamic information of the modal frequency is indirectly obtained by filtering out the mode shape by low pass filter. Then the Hilbert transform is used to process the dynamic component information to receive the instantaneous frequency which is regarded as the damage index. When the structure is damaged, the first instantaneous frequency at the damaged location changes abruptly. To demonstrate the effectiveness and performance of the proposed method, numerical simulations and experimental validations on beam bridge models are conducted. The results show that the method is free from the requirement of the information of bridges under the undamaged state, and only limited sensors are needed to locate the damage.

Exploration of Damage Identification Method for a Large-Span Timber Lattice Shell Structure in Taiyuan Botanical Garden based on Structural Health Monitoring.

Large-span spatial lattice structures generally have characteristics such as incomplete modal information, high modal density, and high degrees of freedom. To address the problem of misjudgment in the damage detection of large-span spatial structures caused by these characteristics, this paper proposed a damage identification method based on time series models. Firstly, the order of the autoregressive moving average (ARMA) model was selected based on the Akaike information criterion (AIC). Then, the long autoregressive method was used to estimate the parameters of the ARMA model and extract the residual sequence of the autocorrelation part of the model. Furthermore, principal component analysis (PCA) was introduced to reduce the dimensionality of the model while retaining the characteristic values. Finally, the Mahalanobis distance (MD) was used to construct the damage sensitive feature (DSF). The dome of Taiyuan Botanical Garden in China is one of the largest non-triangular timber lattice shells worldwide. Relying on the structural health monitoring (SHM) project of this structure, this paper verified the effectiveness of the damage identification model through numerical simulation and determined the damage degree of the dome structure through SHM measurement data. The results demonstrated that the proposed damage identification method can effectively identify the damage of large-span timber lattice structures, locate the damage position, and estimate the degree of damage. The constructed DSF had relatively strong robustness to small damage and environmental noise and has practical application value for SHM in engineering.

Barely visible impact damage detection in composite structures using deep learning networks with varying complexities

Visual inspection is one of the most common non-destructive testing (NDT) methods that offers a fast evaluation of surface damage in aerospace composite structures. However, it is highly dependent on human-related factors and may not detect barely visible impact damage (BVID). In this research, low velocity impact tests with different energy levels are conducted on two groups of composite panels, namely ‘reference’ and ‘sensor-integrated’ samples. Then, the results of impact tests, together with C-scan and visual inspection images, are analysed to define the BVID range and create an original image dataset. Next, four different deep learning models are trained, validated and tested to capture the BVID only from the images of the impacted and non-impacted surfaces. The results show that all four networks can learn and detect BVID quite well, and the sensor-integrated samples reduce the training time and improve the accuracy of deep learning models. ResNet outperforms other networks with the highest accuracy of 96.2% and 98.36% on the back-face of reference and sensor-integrated samples, respectively. The proposed damage recognition method can act as a fast, inexpensive and accurate structural health monitoring tool for composite structures in real-life applications.

Template mass and stiffness for vibration analysis and damage detection of plane structures

This paper investigates he vibration properties of template plane elements in both themes of optimum mass matrix formulation and damage identification. To find the optimum mass matrices for the elements, the classical and unconventional mass formulations for the plane problems are explored. The optimum mass is obtained by the template approach, along with an optimization procedure, based on the perturbation theory and sensitivity analysis. The efficiencies of the proposed optimum mass are investigated through numerical examples for some plane structures. Those pairs of the optimum mass and stiffness, which yield more accurate results in the vibration analysis, are utilized for damage identification. A novel damage localization index based on the in-plane transverse vibrations is proposed and applied successfully in a new two-stage damage identification procedure for planar structures. By using Test Analysis-Model correlation (TAM) and the element damage equations; authors' algorithm overcomes the challenges of limitations in measuring all the degrees of freedoms. Furthermore, the proposed algorithm has the potential to diagnose the false alarms and prevents the ill-condition problem in damage identification. The abilities of the suggested method are demonstrated through 11 different damage scenarios and the presence of noise.

Subspace-based modal identification and uncertainty quantification from video image flows

Vibration measurements for structural health monitoring (SHM) by operational modal analysis (OMA) are classically obtained from sensors that are embedded in or physically attached to the monitored structure, like accelerometers or strain gauges. However, the setup time of these sensors and their restricted number and space coverage limit their monitoring capabilities. Video image-based sensing methods can overcome these shortcomings. With adequate image processing methods, motion signals are extracted from video image flows, which are then processed by system identification methods to estimate modal parameters. In this way, the pixels in selected regions of interest within the images act as a dense network of contactless sensors distributed over the whole structure. In this paper, the efficiency of this video-based approach is demonstrated with laboratory experiments on a cantilever beam, in particular, by evaluating its capability for detecting weak damages mimicked by slight mass modifications. To this end, the steerable filter-based method (ST), that recovers displacements from local phase, is first extended to overcome its motion limitation of one pixel size. Then, the performance of the improved motion extraction method is compared with two other well-established methods in the context of OMA, where natural frequencies, damping ratios and mode shapes with high spatial resolution are estimated together with their uncertainty bounds using covariance-driven subspace identification. The compared methods are evaluated with the help of reference laser displacement measurements as well as a finite element model of the beam, revealing differences in the accuracy of the estimated mode shapes depending on the chosen method for motion extraction. Finally, aiming to investigate early structural damage detection, experiments are carried out under small structural changes and the results are compared to a reference state with the help of estimated uncertainties. Small but statistically significant changes in the modal parameters are detected, showing the potential of the vision based framework for SHM.

Optimization-based damage detection in composite structures using incomplete measurements

Incomplete data poses challenges in accurately assessing structural health and detecting damage. It limits the ability to capture the complete behavior and response of the structure, which may hinder the identification and localization of potential damage or anomalies. Addressing the issue of incomplete data requires developing strategies and algorithms that can effectively handle missing or limited measurements. Using incomplete and noisy measurements, we propose an optimization-based damage detection method for laminated composite plates with closely-spaced eigenvalues. The proposed method consists of two stages. In the first stage, the most probable defective elements are identified by utilizing condensed mode shapes as incomplete noisy inputs for modal residual vectors. This approach significantly reduces the computational effort for damage estimation. The second stage introduces an objective function based on incomplete and noisy Condensed Frequency Response Functions (CFRFs). To optimize the damage quantification, the Improved Particle Swarm Optimization (IPSO) algorithm is employed to minimize errors in the proposed objective function based on CFRFs of damaged and intact laminates. The proposed method is demonstrated on laminated composite plates with different lamination schemes, ply orientations, and multiple damaged elements in different damage scenarios. By evaluating the method on numerical results and comparing it with previous studies, its superiority is demonstrated. Furthermore, the proposed method exhibits robustness to changes in mass distribution in the system investigated by retrofitting extra masses to the plate structures that lead to worsening closely-spaced eigenvalues.

One-dimensional convolutional neural network for damage detection of structures using time series data

One-dimensional convolutional neural network for damage detection of structures using time series data

Subspace features and statistical indicators for neural network-based damage detection

The late opportunities prompted by artificial intelligence have motivated the current research about structural damage detection strategies based on damage-sensitive subspace-based indicators (DI). Precisely, three different methodologies (A), (B), and (C) are discussed for multiclass damage classification with a multi-layer perceptron (MLP) network. Specifically, the network’s inputs combine vibration response statistics with subspace-based features. Method (A) relies on statistical features only, whereas method (B) also considers the most informative subspace-based DI, retrieved from an empirical sensitivity analysis. Finally, method (C) provides a new perspective by overcoming the arbitrary choice of parameters affecting the subspace-based DIs computation. These three methods are tested on a numerical benchmark problem, and the results emphasize the last approach as the most promising methodology. For the sake of further validation purposes, the three methods have been finally tested on an experimental steel I-beam setup, evidencing the effectiveness of informative subspace-based DIs.

Multifunctional applications of topological valley-locked elastic waves

Topological valley metamaterial in classical wave systems has shown great potential for manipulating electromagnetic, acoustic, and elastic waves due to the defect-immune and lossless energy properties. The application prospects of topological valley material on signal processing, structural damage detection, and energy harvesting are worth exploring further. In this study, we combine the valley Hall effect and the tunneling effect of local resonant metamaterials to reveal some novel coupling transport phenomena in elastic systems. Some potential applications are explored experimentally and numerically. First, we implement a low-frequency (<1 kHz) valley-locked waveguide in heterostructures based on the topological valley edge state of the local resonant metamaterial. The robustness of the topological valley-locked waveguide under wide valley-locked layers, impurities, disorder, and bends is verified. Next, we introduce tunneling phenomena into the topological valley-locked waveguide, enabling the realization of an on-off controllable heterojunction by sandwiching crystals with different Dirac cones. The difference between the transmission characteristics of valley-locked waveguides under different tunneling layers (potential barrier and well) is revealed. Finally, we explore potential applications such as signal splitters, energy concentrators, and logical gates, both numerically and experimentally. Beyond the presented applications, this research has promising application prospects for vibration signal processing and high-performance energy harvesting.

Modal sensitivity analysis of acoustic metamaterials for structural damage detection

Metamaterials have garnered significant research interest over the past few decades, primarily due to their unique properties not found in naturally occurring materials. However, when integrated into operational engineering structures, metamaterials can sustain damage, compromising their extraordinary attributes and potentially leading to structural failure. This work marks a novel advancement in the field of metamaterials research, providing an in-depth analysis into the impact of damage on the structural dynamic properties of finite acoustic metamaterials (AMMs) consisting of periodic locally resonant mass-in-mass units. A modal sensitivity analysis is formulated based on equations of motion of a damaged AMM, along side with associated eigenvalue problem and frequency response functions. The critical role of the internal spring is analytically revealed in determining the effective mass of a damaged unit. To evaluate the effects of damage on AMMs, an extensive numerical investigation is conducted on a finite AMM; an damage index is proposed for measuring the modal deformation of each spring. It is unequivocally demonstrated that frequency response functions, eigenvalues, and mode shapes of damaged AMMs undergo substantial changes at frequencies near bandgaps and these changes diminish as frequencies move away from the bandgaps. This behavior directly corresponds to the frequency-dependent changes in the effective mass of a mass-in-mass unit due to damage. Hence, this work constitutes a leap forward in our understanding of the structural dynamics of damaged metamaterials and offers valuable insights that could facilitate the development of more effective damage identification techniques and the realization of resilient metamaterial-based structures.

Microcrack monitoring and fracture evolution of polyolefin and steel fibre concrete beams using integrated acoustic emission and digital image correlation techniques

The use of polymer and steel fibres in plain concrete appears to be an excellent solution for limiting crack propagation and improving the post-ductility performance of concrete structures. Based on this premise, this study investigated the fracture evolution of polyolefin fibre-reinforced concrete (PFRC) and steel fibre-reinforced concrete (SFRC) specimens through the integrated application of two diagnostic techniques, acoustic emission (AE) and digital image correlation (DIC), under three-point bending tests. Based on the processing of AE signals, different AE statistical parameters such as the cumulative number of hits, amplitude distribution, and some representative analysis methods including the b-value method, Ib-value method, and AE intensity analysis methods were selected to analyse the early detection of cracking and post-cracking behaviour in PFRC vs SFRC specimens during mechanical degradation. Simultaneously, the DIC technique was used to validate the fracture evolution of the AE results. Furthermore, to verify the reliability of the AE and DIC results, the damage localisation and fracture evolution of the PFRC versus SFRC specimens were confirmed by integrating the AE fracture energies and DIC outcomes. The tests and analysed results showed that the addition of steel fibres to plain concrete significantly improved the ability to restrict crack propagation and provided higher post-cracking resistance compared to PFRC specimens owing to their stronger fibre–matrix bonding, effective fibre bridging, and crack-arresting mechanism. The present study indicates that the combined AE and DIC techniques are highly effective for the early detection of damage and ductility performance in fibre-reinforced concrete structures.