Proposed Damage Detection Method Research Articles

Bridge Damage Detection Using Ambient Loads by Bayesian Hypothesis Testing for a Parametric Subspace of an Autoregressive Model

This study explores a change detection method in modal properties to automate and generalize in-service damage detection for vibration-based structural health monitoring of bridges. The noisy conditions caused by ambient loading pose difficulty for in-service damage detection because the load-induced noise often masks the difference in the modal properties. The proposed method directly converts measured time series into a simplified anomaly indicator robust against load-induced noise. This study adopts a vector autoregressive model to represent the vibration of bridges. Bayesian inference produces a posterior probability distribution function of the model parameters. Principal component analysis extracts a subspace comparable to the modal properties in the model parameters. Bayesian hypothesis testing quantifies anomalies in the extracted subspace. The feasibility of the proposed method is assessed with vibration data from field experiments conducted on an actual steel truss bridge. The field experiment includes damage severing the truss members. The modal frequencies and mode shapes estimated from the principal component analysis correspond well to earlier reported results. The proposed damage detection method successfully indicated all damage considered in the experiment.

AI-based condition monitoring on mechanical systems using multibody dynamics models

Real-time monitoring of mechanical systems via vibration measurements allows for detection of faults in them and facilitates their predictive maintenance. Use of Artificial Intelligence (AI) in damage detection can provide an automated means for Condition Monitoring (CM) of the system while also allowing for characterization of the various health states when labeled damaged state training datasets are available. In this work, a novel CM framework using Convolutional Neural Networks (CNNs) is presented for damage detection and identification on mechanical systems, applied on an elevator door rail, using simulated data from a Multibody Dynamics (MBD) model of the physical system. First, an optimal MBD model of the actual structure is constructed by means of black box optimization, using a small number of initial healthy state measurements. Several of the most common nonlinear contact force models are examined before integrating the best suited one to the MBD model. Damaged states of the system are then simulated by modeling various fault mechanisms on the optimal MBD model, allowing for generation of labeled simulated training datasets for the CNNs. Uncertainty is introduced to the models by sampling their key parameters from a Gaussian distribution, thus considering the real system’s inherent uncertainty. The trained CNNs’ robustness and accuracy are then validated by accurately classifying faults on the physical system, proving the proposed damage detection method’s generalization capabilities and highlighting its potential. The proposed methodology may find application in mitigating the problem of data scarcity of damaged state responses which is present in most mechanical systems and can provide a means for more efficient predictive maintenance. Successful implementation of the proposed framework adds to the existing AI based damage detection and identification methodologies by allowing for extension of the method to industrial systems.

A novel structural damage detection method using a hybrid IDE–BP model

Recently, artificial intelligence has been widely applied in the field of structural damage detection. In this study, a back-propagation neural network (BPNN) hybrid with an improved differential evolution algorithm (IDE) is proposed to advance the development of structural damage detection. Generally, most engineering structures’ dynamic characteristics can be directly obtained by modal analysis, and the corresponding vibration features can provide information on structural damage. Thus, based on the results of modal analysis, artificial intelligent methods such as BP neural network find their way to locate the damage and quantify the damage severity. However, the existing research shows that the original BP has insufficient accuracy in structural damage detection, especially cannot effectively deal with complex structural damage. In this study, an improved differential evolution algorithm is proposed, introducing a group sampling strategy and an adaptive mutation strategy. Then, the proposed IDE and BP neural networks are hybridized into IDE–BP to promote the neural network’s data classification and regression ability. Subsequently, based on the IDE–BP model, a method for detecting the location and severity of multi-damage scenarios is established. The decision tree algorithm (DT) is utilized to identify the number of damaged elements. To verify the effectiveness of the proposed damage detection method, the IDE–BP model is applied to the damage detection for a steel pipe and a cantilever plate. To study the anti-noise ability and robustness of the IDE–BP model, the noise and incomplete modes are introduced. Compared with the original BP, SVM, KNN, PSO–BP,​ and DE–BP, the numerical results show that the damage detection method based on the IDE–BP model performs better for all single and multiple damage scenarios.

Wavelet packet energy-based damage detection using guided ultrasonic waves and support vector machine

Substantial work has shown that the modulation of structural damage on probing guided ultrasonic waves can result in wave components corresponding to different frequencies, causing wave energy transfer from central frequency to other frequency bands. To make use of the damage-induced wave energy transfer in different frequency bands, a data-driven method which combines wavelet packet decomposition (WPD), principal component analysis (PCA) and support vector machine (SVM), is proposed in this study for structural damage identification in both metallic and composite materials. Firstly, WPD is employed to decompose the original signal into different frequency bands, based on which the wave energy at each frequency band can be determined. A wave energy distribution vector is constructed according to the energy proportion of each frequency band. Then, PCA is recalled conducting dimensionality reduction for the energy distribution vectors, in order to improve the computational efficiency for subsequent SVM classification. The compressed energy distribution vectors are used as the input to train an SVM-based classifier for identifying structural damage. To validate the proposed WPD-PCA-SVM method, experiments are performed on both aluminum plate and glass fiber reinforced polymer (GFRP) laminate. According to the experimental results, the embryonic fatigue crack in the (aluminum) plate and the anomaly in the GFRP laminate can be identified by the proposed method, with a detection accuracy of 92.86% for aluminum plate and 95.45% for GFRP laminate, respectively, demonstrating the effectiveness of the proposed method for damage detection in both metallic and composite materials.

Guided Wave-Based Damage Detection of Square Steel Tubes Utilizing Structure Symmetry

Square steel tube, which is widely used in civil engineering, can suffer from a wide variety of damages and aging defects, thus its nondestructive testing (NDT) has attracted wide attention. This work proposes an ultrasonic guided wave (UGW)-based damage detection method for square steel tubes using structure symmetry. Firstly, the dispersion characteristics of square steel tubes are obtained using the semi-analytical finite element (SAFE) method, after which the optimal guided wave modes for damage located on the long and short edges of the steel tube are selected by modal analysis. Then, using the symmetry of the square steel tube’s section, the symmetric layout scheme of the transmitters and the receivers is designed; on this basis, a signal processing strategy for damage detection is proposed by subtracting the receiving signals obtained from symmetric positions. Finally, the effectiveness of the proposed damage detection method is verified by numerical simulations and laboratory experiments. The results show that the proposed method has good inspection accuracy for crack and hole damages on both the long and short edges of square steel tube because the dispersion effect and clutters can be reduced utilizing structure symmetry. It is worth mentioning that the inspection effect for corner holes can be enhanced by modifying the position of the receivers.

A New Electrical Resistance Measurement Method of Carbon Fiber-Reinforced Polymer Tubes Based on C⁴D Technique

This work introduces the capacitively coupled contactless conductivity detection ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\text{C}^{{4}}\text{D}$ </tex-math></inline-formula> ) technique into nondestructive testing (NDT) of carbon fiber-reinforced polymer (CFRP) and proposes a new resistance measurement method of CFRP tubes. The proposed method does not require direct contact between the electrodes and the carbon fiber. Therefore, the electrical resistance of CFRP can be measured without the treatments required by traditional electrical methods that may lead to surface or structural damage. The equivalent circuit model of the new method is established, and the challenges are analyzed. Simulation and experiment are conducted with woven fabric CFRP tubes to verify the feasibility of the proposed method, including the range of workable frequency, the effectiveness of the circuit model, and the capability of damage detection. Research results show that a relatively high working frequency is required to accurately measure the interested resistance and sense the resistance change. By selecting a workable frequency, the effectiveness of the circuit model is verified. Research results indicate that the stray capacitance between the electrodes has little effect on resistance measurement, and the capacitor–resistor–capacitor model is still applicable. Research results also verify the feasibility of the proposed method for damage detection. It is able to sense the resistance change resulted from the damage of CFRP, either the conductivity changes in simulation or actual cutting damage in experiment.

An unsupervised learning damage diagnosis method based on virtual impulse response function and time series models

As it contains the inherent properties of the structure and is insensitive to variation of the input, the transmissibility function (TF) is widely used for structural damage detection under ambient excitation. However, in TF-based methods, the selection of the optimal damage sensitivity band often requires access to vibration data under different operating conditions, which is not conducive to unsupervised learning. In order to accommodate unsupervised learning and extract more damage information, this paper proposes a data-driven unsupervised damage detection method based on the virtual impulse response function (VIRF) and time series model (TSM). In this method, the VIRF is used as a damage-sensitive feature (DSF), the TSM is used for low-dimensional feature extraction of the VIRF, and the distance metric of the model coefficients is used to detect the occurrence of anomalies. The study analyses the type of TSM applied to VIRF and uses the parameter adjusted cosine similarity (PACS) as a more accurate statistical distance tool for the model coefficients. The application of two well-known benchmark structure experiments verifies the accuracy and practicality of the proposed method in damage detection under minor operational variations and ambient excitations. The comparative analysis and detection results indicate that the proposed method is advantageous in qualitatively quantifying the extent of the damage. Structural anomalies can be detected in both linear and nonlinear damage scenarios.

Unsupervised structural damage detection based on an improved generative adversarial network and cloud model

The measurements derived from damaged conditions are difficult to acquire in actual structures, which limits the applicability of supervised damage detection methods. In this study, an unsupervised damage detection method that leverages an improved generative adversarial network (IGAN) and cloud model (CM) is proposed. This method only needs the data in the healthy state of the structure for model training, which can solve the above problems. Firstly, an IGAN model is established, which uses the encoder-decoder-encoder generative network to encode, reconstruct and re-encode the structural response in healthy state. The essential features hidden in complex data are learned by minimizing the distance between the real response and the reconstructed response, and the distance between the latent vectors obtained after two different encodings. During the test phase, when unknown state data is input into the model, the differences of the latent features of the two different codes can be used to preliminarily judge whether the structure is damaged. In addition, CM theory is used to further quantify structural damage and solve the problem of damage misjudgment caused by measurement noise, model parameters, feature selection, and other uncertain factors. The effectiveness of the proposed damage detection method is verified by Phase I IASC-ASCE benchmark structure and a bridge health monitoring benchmark model. The results prove that the proposed method can accurately detect the damage that has occurred.

A detection method integrating modal deflection curvature difference and natural frequency for structural stiffness degradation

A detection method integrating modal deflection and natural frequency is proposed to locate and quantify the damage of the beam. A damage identification index using modal deflection curvature difference is developed to locate damage. Note that the modal deflection is derived from the flexibility matrix and virtual force. A damage quantification equation is established by the transfer matrix method, which includes the parameters of natural frequencies, damage coefficients and damage location. Since the damage location parameters are determined by the proposed damage identification index, the damage coefficients of the damaged segments can be determined from the natural frequencies according to the damage quantification equation. The proposed damage detection method is verified numerically and experimentally. The results indicate that the proposed method can well locate and quantify the local damage of beams, especially for the beams with non-uniform damage degradation. The proposed damage localization index is more appropriate to locate damage than the index using modal flexibility curvature (MFC) or modal flexibility curvature matrix (MFCM).

Acoustic-based method for identifying surface damage to wind turbine blades by using a convolutional neural network

Damage to the surface of the blades of a large wind turbine may lead to catastrophic blade failure. Although numerous methods have been proposed for detecting surface damage to wind turbine blades, many of them involve laboratory tests because of the difficulty in acquiring data from a commercial wind farm. This lack of data variety is an obstacle to the development of machine learning approaches for identifying the aforementioned damage. Therefore, we developed a damage detection method for wind turbine blade surfaces that is based on the physical correlation between surface conditions and acoustic signals of operating wind turbines under realistic environmental conditions. In the preprocessing stage of the aforementioned method, the short-time Fourier transform and smoothing techniques are used to analyze real-time spectrograms and the rotor speed. The derived spectrogram and rotor speed are then input into a convolutional neural network (CNN) to classify the wind turbine blade surfaces into two classes: turbines with at least one or no damaged blade. The CNN proposed in this paper is a hybrid network containing a masking module and residual classifier. The masking module suppresses redundant information in the spectrogram, and the residual classifier quantifies the difference between the masked spectrogram and a standard spectrogram. The proposed CNN can be easily trained on a small dataset with a few trainable parameters by using the physical characteristics in the residual classifier. The proposed damage detection method was evaluated using the operational noise of commercial wind turbines; according to the results, this method outperformed approaches proposed in previous studies and exhibited an accuracy of 97.11%.

Damage detection of composite laminate structures using VMD of FRF contaminated by high percentage of noise

A novel highly robust-to-noise and closely-situated eigenvalues damage detection method is proposed. The proposed method employs the Variational Mode Decomposition (VMD) algorithm to construct a new set of input signals obtained from the rows of the condensed Frequency Response Function (CFRF) to be used in a sensitivity-based model updating problem. Each row of the FRF matrix is replaced by its Unwrapped Instantaneous Hilbert Phase (UIHP). However, since the signal corresponding to the rows of the CFRF might not exhibit the mono-component property, and thus the UIHP will not be well-defined, VMD is used to obtain a set of constructive mono-component modes for each row, whereby the sum of UIHPs (SUIHP) for that row is obtained. The obtained SUIHPs for all rows of the CFRF are stacked up to obtain a new matrix to be fed into the optimisation problem. The proposed method is tested on a composite laminate plate with different configurations, as an example of structures with closely-situated eigenvalues. The results of the application of highly noisy measurement data for damage detection as well as comparison with two other methods demonstrate the superiority of the proposed method in damage detection of structures with closely-situated eigenvalues using highly noisy input data.

A mode shape sensitivity-based method for damage detection of structures with closely-spaced eigenvalues

A new optimisation problem is proposed to facilitate a fast and reliable damage detection of structures with closely-spaced eigenvalues. The first stage of the proposed method identifies the most probable defective elements resulting in the elimination of healthy members from further investigation. This will further reduce the computational efforts of computing damage indices regarding the defective elements. The second stage of the proposed method exploits the proposed objective function to update the damage indices of the identified defective elements from the first stage. Two truss structures with multiple damaged elements in different damage scenarios are studied where measurements with different levels of noise are used as input to the proposed algorithm. Numerical results and comparison with previous studies demonstrate the superiority of the proposed method in damage detection of structures with closely-spaced eigenvalues.

Damage Identification in Hangers of Through-Arch Bridges Using Static Deflection Difference at the Anchorage Point

This study proposes a new damage identification method for hangers of arch bridges using the static deflection difference at the anchorage point of the hanger and the tie-beam. The relationship between the ratio of cable tension loss and the deflection difference at the anchorage point is studied. For the first time, the deflection difference influence matrix for hanger damage identification is defined. The static deflection change parameter is formulated as a function of the deflection difference influence matrix and the ratio of cable tension loss. The study shows that the percentage of cable tension loss can be obtained from the changes in static deflection at the anchorage point and the deflection difference influence matrix. Therefore, by observing a plot of the ratio of cable tension loss, the damage locations can be identified conveniently. Numerical and laboratory studies show that this method can accurately locate the damaged hanger of through-arch bridges under various scenarios. The proposed damage detection method has a clear theoretical base and is simple to operate, and it is suitable for practical engineering.

Using incomplete FRF measurements for damage detection of structures with closely-spaced eigenvalues

In this paper, a new model updating-based damage detection method based on a modified sensitivity equation is proposed that can tackle the problem of damage detection in structures with closely-spaced eigenvalues. It is known that modal information, such as natural frequencies, in these structures can be of close proximity, making the procedure of damage detection hard if not impossible. The obtained sensitivity equation uses incomplete measurements from frequency response functions (FRFs) to conduct the challenging damage detection of structures with closely-spaced eigenvalues. Although using FRF for damage detection of this kind of structures can offer some advantages over modal data, there are still some challenges that need to be addressed. For instance, it is not possible to have the response of the structure measured in all of its degrees of freedom. As such, the proposed FRF-based model updating method is capable of overcoming these limitations and has the advantage of avoiding modal analysis errors. In order to evaluate the efficiency of the proposed method, two numerical examples of a 144-element three-layered laminated composite plate and a 120-element three-dimensional truss structure, as examples of structures with closely-spaced eigenvalues, are studied. The results demonstrate the capability of the proposed method in damage detection of structures with closely-spaced eigenvalues with incomplete measurement data. Moreover, the results of comparison between the proposed method with some other methods demonstrate the superiority of the proposed method in damage detection of structures with closely-spaced eigenvalues using noisy incomplete FRF data.

Using measured rotation on a beam to detect changes in its structural condition

ABSTRACT A recent survey of Europe’s highway infrastructure has concluded that almost half of Europe’s bridges are nearing the end of their design live. Work in the wider Structural Health Monitoring sector is aiming to develop reliable and cost-effective methods for verifying condition, remaining service life and safety of ageing structures. Most bridge condition assessment methods are based on deflection, acceleration or strain measurements. This paper looks at the possibility of using rotation measurements as a main parameter to identify damage. This study looks at numerical analyses of a moving point load on a one-dimensional bridge model to provide the theoretical basis of the proposed damage detection method. It is shown that when local damage occurs, even when it is remote from a sensor location, it results in an increase in the magnitude of rotation measurements. This study looks at how best to exploit this fact for damage detection. A number of damage scenarios, sensor locations, and load arrangements are investigated in this study and their influence on the ability of the algorithm to detect damage are reported.

Structural damage detection of space frame structures with semi-rigid connections

The accurate modeling of connection joints for civil engineering structures is challenging. The conventional approaches generally assume that the joints are either pinned or rigid, while real joints are semi-rigid in practice. Most existing damage detection methods do not consider the semi-rigidity of the joints, and thus may lead to false or inaccurate identification results. Moreover, the joints are more vulnerable to damage compared to other members of the structure. However, only limited studies have focused on the detection of the joint damage, and few have studied the damage in structural members and joints simultaneously for space structures. In this regard, a two-step damage detection method for space frame structures with semi-rigid connections is developed in this study. Rotational springs with some flexibility are used to model the mechanical behavior of the joints, which enables identification of damage in the bar elements as well as in joints. The first step determines the existence and location of the damaged elements using the l1 regularization technique. In the second step, the damage locations and severities are then identified within the suspicious damaged elements only that are detected in the first step. This process progressively reduces the number of unknown parameters and thus alleviates the instability and divergence problems to improve the damage identification accuracy. An experimental 3D space structure is utilized to demonstrate the effectiveness of the proposed damage detection method. The results show that damage in structural elements and joints can be correctly located and quantified progressively. The numerical investigation shows that the sensitivity of the modal properties with respect to joint damage is much smaller than to damage in bar elements.

A baseline-free damage detection method using VBI incomplete measurement data

A novel baseline-free method for damage detection of vehicle-bridge interaction (VBI) systems is proposed. The proposed method is physics-based, in contrast to many prevailing approaches, which are purely data-based techniques. It uses incomplete measurement data by incorporating the static condensation transformation matrix into the equations to obtain the final formulas. However, the static condensation of the damaged beam is not known a priori. Therefore, it is shown analytically that the static condensation transformation matrix of the undamaged beam can be used instead of the one corresponding to the damaged beam. This has been confirmed through numerical simulations for different boundary conditions of the beam. Various factors are studied numerically in order to demonstrate the robustness of the proposed method, including road roughness, boundary conditions, variable moving mass velocity and measurement noise. The results demonstrate the capability of the proposed method in damage detection of beam-type structures subjected to a moving mass in the presence of 5% noise. It has also been shown that averaging the results obtained from noisy data collected through several repetitions of the experiment can improve the final prediction of the location and severity of damage.

Vibration based damage detection in hyperbolic cooling towers using coupled method

Hyperbolic cooling towers are high rise reinforced concrete structures that are built to last a long time. Their primary task is to remove the heat from the wastewater and transfer it to the atmosphere using the process of evaporation. Hence these structures are subjected to large temperature variations. These along with material degradation, environmental effects, and random actions can initiate damage in these structures. Such damage can go undetected as a significant part of a hyperbolic cooling tower is neither easily visible nor accessible. Determining the onset of damage and carrying out appropriate retrofitting is important to prevent the failure or collapse of these large structures. Relatively few studies have been carried out on damage detection in cooling towers. This paper develops and applies a coupled method based on vibration characteristics using the absolute change in mode shape curvature to detect and locate damage in hyperbolic cooling towers. The coupled method considers few vertical cross-sections to first determine the height of the damage and then its position on the horizontal cross-section at that height of the axisymmetric cooling tower. The proposed damage detection method and modelling techniques are validated using data from experimental testing of laboratory-scale cooling tower models. Effects of noise and limited number of data points are considered to represent practical circumstances. Results confirm that the proposed method is very effective in detecting and locating damage in hyperbolic cooling towers under a range of damage scenarios using the first few global vibration modes. Findings of this paper will contribute towards enhanced safety of these large structures and help to prevent their failure.

A New Damage Detection Method for Special-Shaped Steel Arch Bridges Based on Fractal Theory and the Model Updating Technique

In this paper, an innovative two-level damage detection method applicable to real-world online structural health monitoring (SHM) systems is proposed for in-service large steel arch bridges. The method consists of Level 1 damage detection practice that includes strain data acquisition and damage location using the damage index based on the fractal theory, and Level 2 damage detection practice that includes acceleration sample acquisitions and dynamic model updating to quantify the damage. A numerical case study of the Yingzhou bridge based on various damage cases demonstrated the effectiveness of the proposed damage detection method. It is revealed that Level 1 damage detection is sufficiently robust against the standard measurement noise and normal temperature variations. The study results also indicated that the accuracy of Level 2 damage detection largely depends on whether the initial structure imperfections are taken into account, and whether the utilized model updating method is effective under model errors.

Structural damage detection and localization using decision tree ensemble and vibration data

AbstractThis paper explores the capabilities of decision tree ensembles (DTEs) for detecting and localizing damage in structural health monitoring (SHM). Unlike research on many other learning models, the goal of this study is to identify damage with a localization accuracy down to the single structural element, rather than limiting the evaluation to the story scale. The SHM methodology herein discussed, denoted as D2‐DTE, is based on decision trees ensemble and belongs to the class of vibration‐based approaches, being the health assessment of the structure obtained by analyzing dynamic properties of the structural system, namely, mode shapes and natural frequencies. The proposed damage detection method is validated for three different test cases, including both numerical simulations and experimentally recorded data, which consider a wide array of damage configurations, including single and multiple damages; different damage types and severities; and the presence of random noise levels associated with dynamic properties acquisition. The performances of the D2‐DTE are evaluated in terms of accuracy, confidence of probabilistic predictions, and measurements of physical distances in localization errors. Additionally, two of the investigated test cases are based on available benchmarks, thus allowing a direct comparison with a state‐of‐the‐art methodology. This comparative analysis evidences competitive performances of the DTE learning method.