Crack Identification In Structures Research Articles

Automated crack identification in structures using acoustic waveforms and deep learning

Structural elements undergo multiple levels of damage at various locations due to environments and critical loading conditions. The level of damage and its location can be predicted using acoustic emission (AE) waveforms that are captured from the generation of inherent microcracks. Existing AE methods are reliant on the feature selection of the captured waveforms and may be subjective in nature. To automate this process, this paper proposes a deep-learning model to predict the damage severity and its expected location using AE waveforms. The model is based on a densely connected convolutional neural network (CNN) that offers superior feature extraction and minimal training data requirements. Time-domain AE waveforms are used as inputs of the proposed model to automate the process of predicting the severity of damage and identifying the expected location of the damage in structural elements. The proposed approach is validated using AE data collected from a concrete beam and a wooden beam and plate. The results show the capability of the proposed method for predicting the level of damage with an accuracy range of 92-95% and identifying the approximate location of damage with 90-100% accuracy. Thus, the proposed method serves as a robust technique for damage severity prediction and localization in civil structures.

Improved acoustic source localization method for crack identification in structures

Acoustic Emission (AE) monitoring is considered one of the popular non-destructive testing (NDT) methodologies that have been used to predict and identify the location of damage in critical civil infrastructure. In this paper, an improved AE crack localization method is proposed by integrating the empirical mode decomposition (EMD)-based signal decomposition method with a source localization model. Unlike the conventional AE method, the proposed method is free of the use of bulk AE parameters such as counts, rise time, signal strength, and energy. First, EMD is used to minimize the presence of noise in the recorded AE waveforms and extract the key AE components. Then, key AE events are located using the source localization model to localize the crack in concrete structures. The performance of the proposed method is validated experimentally on small and large-scale concrete beams, where the damage is induced using progressive static load testing. In particular, the large-scale beams are designed for flexural and shear mode failure to evaluate the performance of the proposed method under various types of damage. Finally, the results of the proposed method are compared with the traditional method that uses raw AE waveforms and the method that uses bandpass-filtered AE waveforms. The results show higher crack location accuracy of the proposed method than the other methods, which makes it a suitable approach as a crack localization technology.

Autonomous Identification of Bridge Concrete Cracks Using Unmanned Aircraft Images and Improved Lightweight Deep Convolutional Networks

The identification of the development of structural defects is an important part of bridge structure damage diagnosis, and cracks are considered the most typical and highly dangerous structural disease. However, existing deep learning‐based methods are mostly aimed at the scene of concrete cracks, while they rarely focus on designing network architectures to improve the vision‐based model performance from the perspective of unmanned aircraft system (UAS) inspection, which leads to a lack of specificity. Because of this, this study proposes a novel lightweight deep convolutional neural network‐based crack pixel‐level segmentation network for UAS‐based inspection scenes. Firstly, the classical encoder‐decoder architecture UNET is utilized as the base model for bridge structural crack identification, and the hourglass‐shaped depthwise separable convolution is introduced to replace the traditional convolutional operation in the UNET model to reduce model parameters. Then, a kind of lightweight and efficient channel attention module is used to improve model feature fuzzy ability and segmentation accuracy. We conducted a series of experiments on bridge structural crack detection tasks by utilizing a long‐span bridge as the research item. The experimental results show that the constructed method achieves an effective balance between reasoning accuracy and efficiency with the value of 97.62% precision, 97.23% recall, 97.42% accuracy, and 93.25% IOU on the bridge concrete crack datasets, which are significantly higher than those of other state‐of‐the‐art baseline methods. It can be inferred that the application of hourglass‐shaped depth‐separable volumes can actively reduce basic model parameters. Moreover, the lightweight and efficient attention modules can achieve local cross‐channel interaction without dimensionality reduction and improve the network segmentation performance.

Study on Crack Identification with Responses Modulated by Nonlinear Energy Sink

Owing to its vibration reduction and indirect signal amplification characteristics, a nonlinear energy sink (NES) is used for crack identification of structures. The equation of a cracked simply supported beam with an NES is derived and the influences of the crack properties on the dynamics of the system are investigated. A pixel method is applied to estimate phase diagram areas and crack parameters are identified based on the geometric features of the phase diagram. It is shown that the phase diagram of the NES varies significantly with crack parameters. This is because of the higher amplitudes generated by the absorbing NES with respect to the primary structure. In addition, the NES can reduce the amplitude of the cracked beam without hindering crack identification. Moreover, through introducing the NES, the system can produce a strongly modulated response, which can be utilized for convenient crack detection. Two sets of crack identification indices are defined for the system, generating a single-period response and a strongly modulated response. Contour maps of identification indices based on the area of phase diagrams are obtained by varying crack location and depth. The crack parameters are identified through the intersection of contour lines.

Pearson Correlation and Discrete Wavelet Transform for Crack Identification in Steel Beams

Discrete wavelet transform is a useful means for crack identification of beam structures. However, its accuracy is severely dependent on the selecting mother wavelet and vanishing moments, which raises a significant challenge in practical structural crack identification. In this paper, a novel approach is introduced for structural health monitoring of beams to fix this challenge. The approach is based on the combination of statistical characteristics of vibrational mode shapes of the beam structures and their discrete wavelet transforms. First, this paper suggests using regression statistics between intact and damaged modes to monitor the health of beam structures. Then, it suggests extracting quasi-Pearson-based mode shape index of the beam structures to use them as an original signal in discrete wavelet transforms. Findings show that the proposed approach has several advantages compared with the conventional mode shape signal processing by the discrete wavelet transforms and significantly improves damage detection’s accuracy.

Structural Crack Identification Based on the Variational Mode Decomposition

In order to enrich the method of bridge damage detection and further improve the accuracy of bridge damage identification, a detection method of simply supported beam with crack under dynamic load is proposed, which is not based on the complete finite element model. Under the premise of not blocking traffic, the method only needs to analyze and deal with the acceleration response of simply supported beam span, which reduces the loading, unloading and maintenance of sensors in practical engineering. At the same time, based on the model, an analytical formula of acceleration in the mid-span of simply supported cracked beam is derived. Supported by theoretical derivation, instantaneous energy and mean energy difference are constructed by using variational mode decomposition and Hilbert transformation, which can effectively identify small cracks with 5% crack depth. Based on this, the influences of different wheel load, environmental noise and damage degree on detection results are studied. The results show that: ① The instantaneous frequency has a better recognition effect on crack location; ② The mean energy difference is sensitive to different crack depth ratio δ and wheel load; ③ This method has strong noise robustness.

Automated site-specific assessment of steel structures through integrating machine learning and fracture mechanics

Despite the recent advances in using machine learning for crack identification of structures, the usefulness of artificial intelligence in guiding repair strategies is still limited due to the lack of information on the mechanism of crack formation and propagation. In this context, this paper develops an automated workflow to assess the fracture mechanism of steel structures directly from inspection images. Firstly, a U-Net is developed to find the pixel-location of a crack. Then, a Gaussian Process surrogate model is trained to predict stress intensity factors, K (indicator for fracture mechanism), for cracks with various dimensions using data from finite element simulations. In the end, by taking the dimensions of identified cracks as input, the trained Gaussian Process is used to estimate the K factors. The proposed workflow empowers an inspector to easily identify steel cracks, predict their propagation from a collection of raw images, and make rapid repair decisions on-site.

Crack identification with combined numerical simulation-intelligent optimization algorithms: a brief summary

Crack identification in structures is a typical inverse analysis problem, which is very crucial for the reliability evaluation of various structures. In recent years, the rapid development of numerical technologies and artificial intelligence algorithms has provided a new way for crack detection. Numerical methods are used as the forward analysis tools to solve crack problems, while intelligent optimization approaches are applied to identify crack geometries based on the data collected by forward modeling. In this paper, the research status of crack identification, with the combined typical computational tools and well-known intelligent optimization schemes, is briefly summarized.

Breathing Crack Localization in Structures Based on Principal Component Analysis of Forced Vibration Responses

The identification of a breathing crack is a highly challenging inverse problem in structural health monitoring. A novel output-only damage diagnostic technique based on Principal Component Analysis (PCA) is proposed for breathing crack identification in structures excited by harmonic excitation. The proposed approach basically utilizes the residues obtained from PCA of the forced acceleration-time history responses of the structure for breathing crack identification. In this approach, the traditional single-tone, bitone and as well as multi-tone harmonic excitations are considered as input to the structure while exploring the residues of the responses for breathing crack identification. A new Damage Localization Index (DLI) based on the Fourier spectrum amplitudes of the nonlinear sensitive features (i.e. buried in residues), measured at varied locations spatially across the structure is proposed for breathing crack localization. The robustness and effectiveness of the proposed PCA-based breathing crack localization approach is verified through both numerical and experimental studies.

Structural Crack Detection Using DPP-BOTDA and Crack-Induced Features of the Brillouin Gain Spectrum

Structural damage generally initiates in the form of structural cracks. Thus, developing efficient crack detection techniques is of great importance for the structural health monitoring. In this paper, a new crack identification method is proposed, which is based on the differential pulse-width pair Brillouin optical time domain analysis (DPP-BOTDA) technology and the irregular features of Brillouin gain spectrum (BGS) in the fiber due to structural cracks. The proposed method provides a new way to detect and quantify structural cracks without knowing the strain in the structure. First, the working mechanism of DPP-BOTDA is introduced to illustrate the reason that the DPP-BOTDA, compared to traditional BOTDA technique, can significantly improve the spatial resolution of distributed strain sensing, which is critical for structural crack detection. Then, the BGSs in the fiber with the presence of structural cracks, measured by the DPP-BOTDA, are numerically simulated, from which the crack-induced irregular features of the BGS are summarized. Based these irregular features, new structural crack detection and quantification methods are proposed, which are found to be independent of structural stain. Finally, an experiment is conducted on a simple supported reinforced concrete (RC) beam. The results demonstrate that by using the BGS measured by the DPP-BOTDA, the proposed structural crack identification method successfully detects the occurrence of structural cracks and relatively accurately predicts the crack widths.

Structural crack detection using deep learning–based fully convolutional networks

Cracks are a potential threat to the safety and endurance of civil infrastructures, and therefore, careful and regular structural crack inspection is needed during their long-term service periods. Many image-processing approaches have been developed for structural crack detection. However, like traditional edge detection algorithms, these methods are easily disturbed by the environmental effect. Convolutional neural networks are newly developed methods and have excellent performances in the image-classification tasks. This study proposes a fully convolutional network called Ci-Net for structural crack identification. Pixel-level labeled image training data are obtained from the online data set. Four indices are adopted to evaluate the performance of the trained Ci-Net. Crack images from an indoor concrete beam test are adopted for validation of its structural crack recognition capacity. The recognition results are also compared with those obtained by the edge detection methods. It indicates that Ci-Net exhibits a better performance over the edge detection methods in structural damage detection.

Bayesian optimal sensor placement for crack identification in structures using strain measurements

A Bayesian framework is presented for finding the optimal locations of strain sensors in a plate with a crack with the goal of identifying the crack properties, such as crack location, size, and orientation. Sensor grids of different type and size are considered. The Bayesian optimal sensor placement framework is rooted in information theory, and the optimal grid is the one which maximizes the expected information gain (Kullback–Liebler divergence) between the prior and posterior probability density functions of the crack parameters. The uncertainty in the crack parameters is accounted for naturally within the Bayesian framework through the prior probability density functions. The framework is demonstrated for a thin plate with crack, subjected to static loading. A finite element model is used to simulate the strain distributions in the plate given the crack properties. To verify the effectiveness of the proposed optimal sensor placement methodology, the estimated optimal sensor grids are used to perform Bayesian crack identification using simulated data. Parametric analyses are carried out giving emphasis on the effect of the number of sensors, grid type, and experimental data noise levels in the identification results.

A comparative study on crack identification of structures from the changes in natural frequencies using GA and PSO

Purpose – The early detection of cracks, corrosion and structural failure in aging structures is one of the major challenges in the civil, mechanical and aircraft industries. Common inspection techniques are time consuming and hence can have strong economic implications due to downtime. The paper aims to discuss these issues. Design/methodology/approach – As a result, during the past decade a number of methodologies have been proposed for detecting crack in structure based on variations in the structure's dynamic characteristics. This work showcases the efficacy of particle swarm optimization (PSO) and genetic algorithm (GA) in damage assessment of structures. Findings – Efficiency of these tools has been tested on structures like beam, plane and space truss. The results show the effectiveness of PSO in crack identification and the possibility of implementing it in a real-time structural health monitoring system for aircraft and civil structures. Originality/value – The methodology presented establishes the PSO as robust and competent tool over GA for crack identification using changes in natural frequencies.

Three-steps-meshing based multiple crack identification for structures and its experimental studies

Multiple crack identification plays an important role in vibration-based crack identification of structures. Traditional crack detection method of single crack is difficult to be used in multiple crack diagnosis. A three-step-meshing method for the multiple cracks identification in structures is presented. Firstly, the changes in natural frequency of a structure with various crack locations and depth are accurately obtained by means of wavelet finite element method, and then the damage coefficient method is used to determine the number and the region of cracks. Secondly, different regions in the cracked structure are divided into meshes with different scales, and then the small unit containing cracks in the damaged area is gradually located by iterative computation. Lastly, by finding the points of intersection of three frequency contour lines in the small unit, the crack location and depth are identified. In order to verify the effectiveness of the presented method, a multiple cracks identification experiment is carried out. The diagnostic tests on a cantilever beam under two working conditions show the accuracy of the proposed method: with a maximum error of crack location identification 2.7% and of depth identification 5.2%. The method is able to detect multiple crack of beam with less subdivision and higher precision, and can be developed as a multiple crack detection approach for complicated structures.

A Bayesian methodology for crack identification in structures using strain measurements

A Bayesian system identification methodology is presented for estimating the crack location, size and orientation in a structure using strain measurements. The Bayesian statistical approach combines information from measured data and analytical or computational models of structural behaviour to predict estimates of the crack characteristics along with the associated uncertainties, taking into account modelling and measurement errors. An optimal sensor location methodology is also proposed to maximise the information that is contained in the measured data for crack identification problems. For this, the most informative, about the condition of the structure, data are obtained by minimising the information entropy measure of the uncertainty in the crack parameter estimates. Both crack identification and optimal sensor location formulations lead to highly non-convex optimisation problems in which multiple local and global optima may exist. A hybrid optimisation method, based on evolutionary strategies and gradient-based techniques, is used to determine the global minima. The effectiveness of the proposed methodologies is illustrated using simulated data from a single crack in a thin plate subjected to static loading.

Identification of Small Crack in Beam Structures Using Anti-Resonant Frequency and Wavelet Analysis

Structural crack identification has been received considerable attention in recent decade. A lot of different techniques like acoustic emission, ultrasonic, or X-ray, etc have been used for structural crack detection. However, it is still difficult to identify the small crack in structures. A new method for identification of small crack in beam structures using the first anti-resonant frequency curve is proposed in this paper. The method makes use of the driving-point mechanical impedance characteristics of beam structures and a simplified rotational spring model to model the edge crack of beam. After the first anti-resonant frequency curve is obtained, signal process based on wavelet transformation will be carried out and the small crack in beam structures can be explored. The proposed method is validated by a numerical example of cracked beam with pinned-pinned or fixed-free boundary. It is concluded that not only the location of beam crack can be determined, but also the extent of crack damage can be identified qualitatively based on the first anti-resonant frequency curve and wavelet analysis.

XFEM‐based crack detection scheme using a genetic algorithm

AbstractA new computational tool is developed for the accurate detection and identification of cracks in structures, to be used in conjunction with non‐destructive testing of specimens. It is based on the solution of an inverse problem. Based on some measurements, typically along part of the boundary of the structure, that describe the response of the structure to vibration in a chosen frequency or a combination of frequencies, the goal is to estimate whether the structure contains a crack, and if so, to find the parameters (location, size, orientation and shape) of the crack that produces a response closest to the given measurement data in some chosen norm. The inverse problem is solved using a genetic algorithm (GA). The GA optimization process requires the solution of a very large amount of forward problems. The latter are solved via the extended finite element method (XFEM). This enables one to employ the same regular mesh for all the forward problems. Performance of the method is demonstrated via a number of numerical examples involving a cracked flat membrane. Various computational aspects of the method are discussed, including the a priori estimation of the ill‐posedness of the crack identification problem. Copyright © 2007 John Wiley & Sons, Ltd.

Crack detection in simply supported beams without baseline modal parameters by stationary wavelet transform

There are significant changes in the vibration responses of cracked structures when the crack depth is significant in comparison to the depth of the structure. This fact enables the identification of cracks in structures from their vibration response data. However when the crack is relatively small, it is difficult to identify the presence of the crack by a mere observation of the vibration response data. A new approach for crack detection in beam-like structures is presented and applied to cracked simply supported beams in this paper. The approach is based on finding the difference between two sets of detail coefficients obtained by the use of the stationary wavelet transform (SWT) of two sets of mode shape data of the beam-like structure. These two sets of mode shape data, which constitute two new signal series, are obtained and reconstructed from the modal displacement data of a cracked simply supported beam. They represent the left half and the modified right half of the modal data of the simply supported beam. SWT is a redundant transform that doubles the number of input samples at each iteration. It provides a more accurate estimate of the variances at each scale and facilitates the identification of salient features in a signal, especially for recognising noise or signal rupture. It is well known that the mode shape of a beam containing a small crack is apparently a single smooth curve like that of an uncracked beam. However, the mode shape of the cracked beam actually exhibits a local peak or discontinuity in the region of damage. Therefore, the mode shape ‘signal’ of a cracked beam can be approximately considered as that of the uncracked beam contaminated by ‘noise’, which consists of response noise and the additional response due to the crack. Thus, the modal data can be decomposed by SWT into a smooth curve, called the approximation coefficient, and a detail coefficient. The difference of the detail coefficients of the two new signal series includes crack information that is useful for damage detection. The modal responses of the damaged simply supported beams used are computed using the finite element method. For real cases, mode shape data are affected by experimental noise. Therefore, mode shape data with a normally distributed random noise are also studied. The results show that the proposed method has great potential in crack detection of beam-like structures as it does not require the modal parameters of an uncracked beam as a baseline for crack detection. The effects of crack size, depth and location, and the effects of sampling interval are examined.

Application of Wavelet Theory for Crack Identification in Structures

This paper presents the first attempt of an application of the wavelet theory for the crack identification of structures. As a case study, we consider the crack identification using the wavelet theory for a simply supported beam with a transverse on-edge nonpropagating open crack. A mathematical model of the cracked beam is derived and the wavelet expressions in the space domain are proposed for solution. For comparison purposes, the simply supported cracked beam is solved using both the eigentheory and the wavelet method of analysis. The results show that crack identification via wavelet analysis is accomplished easily whereas it can hardly be detected by the traditional eigenvalue analysis.

Crack identification in a cantilever beam under uncertain end conditions

Crack identification in structures by changes of their dynamic behavior has been studied in the past, and various methods were developed enabling the calculation of crack location along a beam, by using the variations in the natural frequencies between the initial undamaged state and a later, cracked beam. Application of this procedure to cantilever beams may result in unacceptably large errors, due to changes in clamp rigidity between measurements in the two states. The present research studies the problem of crack identification in a cantilever when clamp rigidity is unknown, and may change with time. An identification method is developed, which requires monitoring of three natural bending frequencies. Crack location may then be found by using a universal curve, i.e. independent of any beam property (geometry or material). The proposed method was verified by numerical simulation and experiment.