Damage Localization Approach Research Articles

Artificial Intelligence-Based Approach for Damage Localization in Ultrasonic Guided Wave-Based Structural Health Monitoring

The continuous monitoring of structural integrity is crucial, as imperceptible damage may appear at any point throughout a structure's lifespan. Several Structural Health Monitoring (SHM) technologies have been developed so far to detect and assess defects in structures. Deep Learning is a common tool for processing data obtained with SHM systems. Although many artificial intelligence-based SHM technologies exist already for damage detection, only a few are focused on damage localization. Existing localization approaches are limited through them being applied on defined simple structures and requiring a large amount of data, which is usually unavailable in practical applications. In this work, measured data from guided ultrasonic wave propagation is used to determine the location of damage in a composite stiffened structure representative of fuselage segments. A novel artificial intelligencebased approach for damage localization is presented and tested with a feed-forward as well as a convolutional neural network. Both architectures show that localization is possible. The accuracy is then analyzed with the probability of localization method and compared to existing non-artificial intelligence-based approaches. These results make it possible to define the minimum damage that can be correctly localized.

Efficacy of modal curvature damage detection in various pre-damage data assumptions and modal identification techniques

The efficacy of modal curvature approach for damage localization is discussed in the paper in the context of input data. Three modal identification methods, i.e., Eigensystem Realization Algorithm (ERA), Natural Excitation Technique with ERA (NExT-ERA) and Covariance Driven Stochastic Subspace Identification (SSI-Cov), and four methods of determining baseline data, i.e., real measurement of the undamaged state, analytical function, Finite Element (FE) model and approximation of current experimental mode shape, are considered. Practical conclusions are formulated based on analysis of two cases. The first is a laboratory beam with a notch and the second is a stonemasonry historic lighthouse with modern restoration in its upper part. The analysis shows that NExT-ERA and SSI-Cov in combination with approximation of current mode shape provide high efficacy in damage localization alongside relatively straightforward determination of baseline data. It proves that the construction of advanced FE models of a structure can be replaced with a much simpler method of baseline data acquisition. Furthermore, the research shows the structural mode shapes identified with ERA may not always indicate the presence of damage.

Enhanced damage segmentation in RC components using pyramid Haar wavelet downsampling and attention U-net

Damage identification in post-earthquake reinforced concrete (RC) structures based on semantic segmentation has been recognized as a promising approach for rapid and non-contact damage localization and quantification. In damage segmentation tasks, damage regions are often set against complex backgrounds, featuring irregular geometric boundaries and intricate textures, posing significant challenges to model segmentation performance. Additionally, the absence of public datasets exacerbates these challenges, hindering advancements in this field. In this paper, a pyramid Haar wavelet downsampling attention UNet (PHA-UNet) semantic segmentation network is proposed, and a database containing 1400 images of damaged RC components (PEDRC-Dataset) with pixel-level annotations is established. In the proposed PHA-UNet, attention mechanisms, multiscale feature fusion, Haar wavelet downsampling, and transfer learning are introduced to address above challenges. Finally, the proposed PHA-UNet is compared with four existing image segmentation architectures on both the Cityspace and the PEDRC-Dataset.

A Hybrid Approach for Damage Detection and Localization on a Thin Metallic Plate

This paper illustrates a hybrid method for identifying structural damage that includes concepts from Physics-Based approaches within a Data-Driven framework. The lessons learned from the use of damage identification methods based on the analysis of modal curvature allows us to develop effective feature engineering in the development of Machine Learning (ML) algorithms. Moreover, the training of the algorithms exploits data provided from a population of damage cases generated by Finite Element (FE) analysis. The above data-analysis pipeline is applied to the identification of damage in a rectangular metallic plate for which measured modal data are available from a dedicated experimental campaign in CNR-INM laboratory using accelerometers. The damage is modelled as a stiffness reduction over a small area being also representative of material thinning due to corrosion. To have meaningful samples of the training database, the modal convergence of the FE model to the real structure is guaranteed by a structural optimization process. Experimentally identified noise, representative of real-life applications, is then added to the FE results before algorithm training. Damage existence and position are determined by a Novelty Detection approach and a Regression Neural Network, respectively. First, damage is identified on new test cases generated by the same FE procedure used for training. Secondly, the trained algorithms are applied to the experimental dataset. Sensitivity analysis on several parameters (number of samples for training, damage severity and noise levels) is numerically carried out to understand the applicability limits of the present methodology.

Utilizing wavelet packet decomposition to locate the damage of composite structures in strong-noise environment

Abstract This paper proposes an approach for damage localization in a strong noise environment based on wavelet packet decomposition. Experiments were conducted to locate damage in composite structures to verify the method. The signal was decomposed by a wavelet packet, and the damage information was identified through wavelet packet energy ratio deviation, defined as the damage factor. The proposed method effectively localizes the damage and directly captures the original signal without the need to preprocess the strong noise signals. Additionally, it bypasses the complex process of extracting reflected signals, demonstrating substantial potential for detecting and identifying composite damage in noisy environments.

Damage localization on reinforced concrete slab structure using electromechanical impedance technique and probability-weighted imaging algorithm

Damage localization plays a key role in structural damage identification and health monitoring. This paper proposed a new damage localization approach using electromechanical impedance/admittance (EMI/EMA) technique integrated with probability-weighted imaging algorithm based on Gaussian distribution. The proposed approach comprises five steps namely measurement of the EMA signature from an array of piezoelectric ceramic lead zirconate titanate (PZT) transducers, computation of statistic damage indicator, establishment of regression model for the indicator, damage visualization via Gaussian distribution and final damage localization. Experimental validation of the approach was conducted on a reinforced concrete (RC) slab with artificial crack damages detected by using an array of surface-bonded PZT patches. Comparative studies on localization accuracy were performed by utilizing root mean square deviation (RMSD), baseline-changeable RMSD namely RMSDk, mean absolute percentage deviation (MAPD) and correlation coefficient (CC). Experimental results demonstrated that the proposed approach could achieve precise localization of concrete crack with error as small as 0 mm, and that using RMSDk index exhibited the highest localization accuracy in average. The proposed approach promoted the capability of the EMI technique in characterization of concrete structural damages.

Internet of things (IoT)-based structural health monitoring of laboratory-scale civil engineering structures

Rapid advances in the Internet of Things (IoT) domain have made it a crucial technology for the real-time structural health monitoring (SHM) of civil engineering infrastructures. The availability of quick and accurate vibration data is essential for SHM, and such data can be obtained through IoT devices mounted on the structures. This study proposes a real-time damage prediction and localization approach using a low-cost "do-it-yourself" wireless sensor node with IoT capabilities for SHM. The proposed sensor node comprised a microcontroller (NODE MCU ESP8266) and a 6-axis accelerometer (MPU6050). The IoT devices track the real-time frequency of the laboratory-scale structure indirectly via measurement of acceleration-time history, and their results are compared with conventional industry-standard accelerometers. Promising results, with a <6% average difference from the conventional accelerometer (difference ranging from 1.3 to 14.3%), provided an innovative SHM for vibration-based real-time SHM using the IoT paradigm. The performance of the proposed methodology was validated numerically and experimentally on two laboratory-scale structures, and the potential of IoT technology for enhancing the efficiency of SHM was demonstrated. The proposed method thus can enable the early detection of damages in infrastructures such as buildings and bridges and thus can reduce the likelihood of accidents via continuous SHM.

Two-Dimensional Damage Localization Using a Piezoelectric Smart Aggregate Approach-Implementation on Arbitrary Shaped Concrete Plates.

This paper presents the application of a hybrid approach for damage localization in concrete plates of arbitrary geometric shapes and a constant thickness. The hybrid algorithm utilizes fast discrete wavelet transformation, energy approach and time of flight criteria for the purpose of the localization of single- and multi-damage problems inside or on the periphery of concrete plates. A brief theoretical background of the hybrid method as well as numerical procedures for modeling the piezoelectric smart aggregate and ultrasonic wave propagation are presented. Experimental and numerical verification of the damage localization were performed on square samples/models with one or two damages and with 16 positions of piezoelectric smart actuator/sensor aggregates. After the verification of the hybrid method, a numerical simulation was performed on models with one or two damages for plates of arbitrary geometric shapes. Based on the obtained results, it was concluded that the proposed method can be applied to damage localization in concrete plates of arbitrary geometric shapes. The presented method and numerical procedure can be further used in research through varying the geometry, number and position of damages as well as the number and position of piezoelectric smart aggregates.

An improved convolutional neural network approach for damage localization in composite materials based on modal analysis

Damage detection of composite materials using modal parameters has limitations in terms of sensitivity to small or localized damage and limited accuracy in damage localization. To address this issue, an enhanced channel attention residual network (ECARNet) damage detection model for composite laminates is proposed. First, finite element analysis is used to obtain training samples, which are processed as two-dimensional data to take full advantage of the convolutional neural network. Then, the residual module uses a multilayer perceptron instead of the traditional convolutional layers to learn the correlation between channels to enhance the generalization ability of the model, and uses the tanh activation function to retain negative information. Finally, a channel focus mechanism is introduced to enable the network to learn key features adaptively. Experimental results on two datasets with different levels of damage demonstrate the superior detection performance of ECARNet, achieving average detection accuracies of 98.13% and 97.94% respectively. A comparison with other methods verifies the effectiveness and reliability of the proposed approach. Furthermore, the effectiveness of the new method is validated on real-world test data.

Perturbation approach for damage localization in beam-type structures: analytical, experimental and numerical exposition

ABSTRACT Structures are usually designed to undergo some yielding, cracking and damage during any catastrophic events like earthquakes. It is necessary to identify these damage locations to avoid the failure of the structure. In this study, a novel method of damage localization for a linear one-dimensional mathematical model of Euler-Bernoulli beam is developed. The damage is modeled as a normalized box-car function. The first-order perturbation in the form of a small change in the stiffness is introduced to the structure. This study employed an effective boxcar filtration technique in damage localization. The proposed formulation shows a distinct peak at the damage location. Further, a two-point roving technique is employed on the experimental model of an overhanging beam under impact loading to check the effectiveness of the proposed localization procedure under real measurement conditions. For its entirety, the finite element model for different end conditions through numerical simulations is also briefly addressed. It is observed that the results obtained from the experimental investigation and the simulation studies are in agreement with the proposed formulation. The proposed methodology does not consider the effect of noise, which can be addressed as the future scope of the present study.

A global-local damage localization and quantification approach in composite structures using ultrasonic guided waves and active infrared thermography

The paper emphasizes an effective quantification of hidden damage in composite structures using ultrasonic guided wave (GW) propagation-based structural health monitoring (SHM) and an artificial neural network (ANN) based active infrared thermography (IRT) analysis. In recent years, there has been increased interest in using a global-local approach for damage localization purposes. The global approach is mainly used in identifying the damage, while the local approach is quantifying. This paper presents a proof-of-study to use such a global-local approach in damage localization and quantification. The main novelties in this paper are the implementation of an improved SHM GW algorithm to localize the damages, a new pixel-based confusion matrix to quantify the size of the damage threshold, and a newly developed IRT-ANN algorithm to validate the damage quantification. From the SHM methodology, it is realized that only three sensors are sufficient to localize the damage, and an ANN- IRT imaging algorithm with only five hidden neurons in quantifying the damage. The robust SHM methods effectively identified, localized, and quantified the different damage dimensions against the non-destructive testing-IRT method in different composite structures.

Unsupervised learning framework for temperature compensated damage identification and localization in ultrasonic guided wave SHM with transfer learning

Damage localization algorithms for ultrasonic guided wave-based structural health monitoring (GW-SHM) typically utilize manually-defined features and supervised machine learning on data collected under various conditions. This scheme has limitations that affect prediction accuracy in practical settings when the model encounters data with a distribution different from that used for training, especially due to variation in environmental factors (e.g., temperature) and types of damages. While deep learning based models that overcome these limitations have been reported in literature, they typically comprise of millions of trainable parameters. As an alternative, we propose an unsupervised approach for temperature-compensated damage identification and localization in GW-SHM systems based on transferring learning from a convolutional auto encoder (TL-CAE). Remarkably, without using signals corresponding to the damages during training (unsupervised), our method demonstrates more accurate damage detection and localization as well as robustness to temperature variations than supervised approaches reported on the publicly available Open Guided Waves (OGW) dataset. Additionally, we have demonstrated reduction in number of trainable parameters using transfer learning (TL) to leverage similarities between time-series among various sensor paths. It is also worth noting that the proposed framework uses raw time-domain signals, without any pre-processing or knowledge of material properties. It should therefore be scalable and adaptable for other materials, structures, damages, and temperature ranges, should more data become available in the future. We present an extensive parametric study to demonstrate feasibility of the proposed method.

Data-driven method for damage localization on soft robotic grippers based on motion dynamics.

Damage detection is one of the critical challenges in operating soft robots in an industrial setting. In repetitive tasks, even a small cut or fatigue can propagate to large damage ceasing the complete operation process. Although research has shown that damage detection can be performed through an embedded sensor network, this approach leads to complicated sensorized systems with additional wiring and equipment, made using complex fabrication processes and often compromising the flexibility of the soft robotic body. Alternatively, in this paper, we proposed a non-invasive approach for damage detection and localization on soft grippers. The essential idea is to track changes in non-linear dynamics of a gripper due to possible damage, where minor changes in material and morphology lead to large differences in the force and torque feedback over time. To test this concept, we developed a classification model based on a bidirectional long short-time memory (biLSTM) network that discovers patterns of dynamics changes in force and torque signals measured at the mounting point. To evaluate this model, we employed a two-fingered Fin Ray gripper and collected data for 43 damage configurations. The experimental results show nearly perfect damage detection accuracy and 97% of its localization. We have also tested the effect of the gripper orientation and the length of time-series data. By shaking the gripper with an optimal roll angle, the localization accuracy can exceed 95% and increase further with additional gripper orientations. The results also show that two periods of the gripper oscillation, i.e., roughly 50 data points, are enough to achieve a reasonable level of damage localization.

A Bayesian approach for damage assessment in welded structures using Lamb-wave surrogate models and minimal sensing

The increasing mechanical and economical demands in modern systems and structures are forcing an inevitable need for joining dissimilar materials, thus creating the challenge of establishing a process to inspect and monitor dissimilar joints. Condition monitoring is a necessity to ensure that the structures are being safely used and to extend their lifetime. Lamb waves (LWs) are ultrasonic guided waves that are widely used for structural health monitoring of mechanical, aerospace, and civil structures. This paper proposes a novel structural health monitoring approach for damage detection, localization, and assessment using a minimal LW sensor-actuator set-up. More specifically, the proposed framework provides damage detection, localization and assessment within a dissimilar-material joint by Bayesian inference of six parameters of damage extent and location. Finite element simulations are used to simulate the measured LWs and generate a dataset required to train artificial neural networks (ANN), acting as surrogate models for LW simulation with reduced computational cost. The ANN-based LW simulations are further used as forward model within an Approximate Bayesian Computation (ABC) framework to provide probabilistic inference of the damage size and position. The results show that damage of different sizes and locations can be successfully identified with a high level of resolution and with quantified uncertainty. The results also show that data fusion for ABC inference using multiple sensor measurements can be possible with improved inference results. However, a precise and robust damage inference can be achieved using a minimal sensing set-up based on one actuator and two sensing points, with consideration of certain levels of measurement noise. These findings imply a considerable reduction of complexity of LW actuator-sensor networks, and overall, they imply a significant reduction of computational resources and cost for damage detection and assessment in structures, thus providing a step forward towards online/onboard monitoring applications.

Bayesian‐optimized unsupervised learning approach for structural damage detection

AbstractStructural health monitoring (SHM) is developing rapidly to fulfill the world's need for resilient and sustainable communities. Due to the current advancements in machine learning and data science, data‐driven SHM is an attractive solution for real‐time damage detection compared to the traditional nondestructive evaluation techniques. However, most widely available data‐driven SHM methods rely on fully or partially simulated data to train the statistical model, and thus require a number of predefined assumptions and parameters, or are not adapted for post‐extreme events damage diagnosis. In this study, we propose a density‐based unsupervised learning approach for structural damage detection and localization. This approach leverages cumulative intensity measures for damage‐sensitive feature extraction for the first time in an unsupervised learning approach. Furthermore, a statistical model construction process is proposed based on kernel density maximum entropy (KDME) and Bayesian optimization. The framework is evaluated in three case studies. The first two involve a numerical three‐story building and a numerical nine‐story asymmetrical building that are both subjected to 100 ground motion excitations while considering environmental variations. The proposed framework is able to detect and localize damage in those case studies with an average accuracy of 92%. The third case study, which contains 44 shake‐table tests of a three‐story frame structure with masonry infill, is used to experimentally validate the proposed framework in damage detection. The three case studies demonstrate the potential and robustness of the proposed Bayesian‐optimized, multivariate KDME novelty detection framework for detecting and localizing structural damage, especially after extreme events.

A new efficient two-stage method for damage localization and quantification in shell structures

The article proposes a new efficient two-stage approach for damage localization and quantification in shell structures using a modal flexibility sensitivity-based damage indicator abbreviated as MFBDI and a recently developed parameter-free optimization algorithm named golden ratio optimization method (GROM). In the first stage, the damage indicator MFBDI is employed to localize possible damage elements in the monitored shell structure. These possible damage elements also help define the search space of optimization problem in the next step. In the second stage, the GROM as a robust optimization solver is implemented to update the finite element (FE) model of the shell structure for refined localization of damage and quantification of its severity. The accuracy and efficiency of the proposed two-stage approach are demonstrated by two numerical simulation examples including a hypar shell and a spherical shell. The simultaneous influences of spatially-incomplete and inaccurate vibration data on damage prediction results are also taken into consideration. The obtained results reveal that the proposed approach can provide an efficient and accurate damage localization and quantification procedure for the studied shell structures.

Damage localization and quantification for beam bridges based on frequency variation of parked vehicle-bridge systems

Frequency variations are normally employed for damage detection as frequencies possess the advantages of the high accuracy and high anti-noise capability. However, the validity and applicability of such methods are restricted due to the limited order of frequencies. A damage localization and quantification approach is proposed for beam bridges by using the frequency variation of parked vehicle-bridge systems. Firstly, the phenomenon of frequency variation of parked vehicle-bridge systems is briefly illustrated. Secondly, a damage location index (DLI) is defined by the frequency change rate (FCR) of the vehicle-bridge systems with the vehicle parked at different positions before and after damage. Then the relationship between the frequency-parameter change rate (FPCR) and damage severity is established via finite element model (FEM) and further employed to formulate a damage equation(s) for severity estimation. The dimension of coefficient matrix in damage equations is reduced to large extent as the damage locations are determined by DLI in advance, which is meaningful in enhancing the effectiveness of severity estimation. Results of numerical and experimental examples with different damage scenarios indicate that the proposed approach can localize and quantify damages with high accuracy and efficiency by using frequency information only. The effects of measurement noise and FEM error are discussed to examine the robustness of the proposed approach. The cases when damage occurs in neighbor elements are analyzed to investigate the feasibility and robustness of the proposed approach.

Robust Baseline-Free Damage Localization by Using Locally Perturbed Dynamic Equilibrium and Data Fusion Technique.

Mode shape-based structural damage identification methods have been widely investigated due to their good performances in damage localization. Nevertheless, the evaluation of mode shapes is severely affected by the measurement noise. Moreover, the conventional mode shape-based damage localization methods are normally proposed based on a certain mode and not effective for multi-damage localization. To tackle these problems, a novel damage localization approach is proposed based on locally perturbed dynamic equilibrium and data fusion approach. The main contributions cover three aspects. Firstly, a joint singular value decomposition technique is proposed to simultaneously decompose several power spectral density transmissibility matrices for robust mode shape estimation, which statistically deals better with the measurement noise than the traditional transmissibility-based methods. Secondly, with the identified mode shapes, an improved pseudo-excitation method is proposed to construct a baseline-free damage localization index by quantifying the locally damage perturbed dynamic equilibrium without the knowledge of material/structural properties. Thirdly, to circumvent the conflicting damage information in different modes and integrate it for robust damage localization, a data fusion scheme is developed, which performs better than the Bayesian fusion approach. Both numerical and experimental studies of cantilever beams with two cracks were conducted to validate the feasibility and effectiveness of the proposed damage localization method. It was found that the proposed method outperforms the traditional transmissibility-based methods in terms of localization accuracy and robustness.

Instant bridge visual inspection using an unmanned aerial vehicle by image capturing and geo-tagging system and deep convolutional neural network

The structural condition of bridges is generally assessed using manual visual inspection. However, this approach consumes labor, time, and capital, and produces subjective results. Therefore, industries today are using automated visual inspection approaches, which quantify and localize damages such as cracks using robots and computer vision. This paper proposes an instant damage identification and localization approach that uses an image capturing and geo-tagging system and deep convolutional neural network for crack detection. The image capturing and geo-tagging allows the geo-tagging of three-dimensional coordinates and camera pose data with bridge inspection images; the deep convolutional neural network is trained for automated crack identification. The damages extracted by the convolutional neural network are instantly transformed into a global bridge damage map, with georeferencing data acquired using the image capturing and geo-tagging. This method is experimentally validated through a lab-scale test on a wall and a field test on a bridge to demonstrate the performance of the instant damage map.

Decomposition Approach for Damage Detection, Localization, and Quantification for a 52-Story Building in Downtown Los Angeles

AbstractAmong the most challenging problems in the field of damage detection and condition assessment in large structures is the ability to reliably detect, locate, and quantify relatively small ch...