A Tensor-Based Structural Damage Identification and Severity Assessment

Development and Validation of a Low-Cost Device for Real-Time Detection of Fatigue Damage of Structures Subjected to Vibrations

This paper presents the development and validation of a low-cost device for real-time detection of fatigue damage of structures subjected to vibrations. The device consists of an hardware and signal processing algorithm to detect and monitor variations in the structural response due to damage accumulation. The effectiveness of the device is demonstrated through experimental validation on a simple Y-shaped specimen subjected to fatigue loading. The results show that the device can accurately detect structural damage and provide real-time feedback on the health status of the structure. The low-cost and easy-to-implement nature of the device makes it promising for use in structural health monitoring applications in various industrial sectors.

Unsupervised Learning Methods for Data-Driven Vibration-Based Structural Health Monitoring: A Review

Structural damage detection using unsupervised learning methods has been a trending topic in the structural health monitoring (SHM) research community during the past decades. In the context of SHM, unsupervised learning methods rely only on data acquired from intact structures for training the statistical models. Consequently, they are often seen as more practical than their supervised counterpart in implementing an early-warning damage detection system in civil structures. In this article, we review publications on data-driven structural health monitoring from the last decade that relies on unsupervised learning methods with a focus on real-world application and practicality. Novelty detection using vibration data is by far the most common approach for unsupervised learning SHM and is, therefore, given more attention in this article. Following a brief introduction, we present the state-of-the-art studies in unsupervised-learning SHM, categorized by the types of used machine-learning methods. We then examine the benchmarks that are commonly used to validate unsupervised-learning SHM methods. We also discuss the main challenges and limitations in the existing literature that make it difficult to translate SHM methods from research to practical applications. Accordingly, we outline the current knowledge gaps and provide recommendations for future directions to assist researchers in developing more reliable SHM methods.

Multi-objective variational autoencoder: an application for smart infrastructure maintenance

Multi-way data analysis has become an essential tool for capturing underlying structures in higher-order data sets where standard two-way analysis techniques often fail to discover the hidden correlations between variables in multi-way data. We propose a multi-objective variational autoencoder (MO-VAE) method for smart infrastructure damage detection and diagnosis in multi-way sensing data based on the reconstruction probability of autoencoder deep neural network (ADNN). Our method fuses data from multiple sensors in one ADNN at which informative features are being extracted and utilized for damage identification. It generates probabilistic anomaly scores to detect damage, asses its severity and further localize it via a new localization layer introduced in the ADNN. We evaluated our method on multi-way laboratory-based and real-life structural datasets in the area of structural health monitoring for damage diagnosis purposes. The data was collected from our deployed data acquisition system on a cable-stayed bridge in Western Sydney, a reinforced concrete cantilever beam which replicates one of the major structural components on the Sydney Harbour Bridge and a laboratory based building structure obtained from Los Alamos National Laboratory (LANL). Experimental results show that the proposed method can accurately detect structural damage. It was also able to estimate the different levels of damage severity, and capture damage locations in an unsupervised aspect. Compared to the state-of-the-art approaches, our proposed method shows better performance in terms of damage detection and localization.

Investigation on Applicability and Limitation of Cosine Similarity-Based Structural Condition Monitoring for Gageocho Offshore Structure

The key to coping with global warming is reconstructing energy governance from carbon-based to sustainable resources. Offshore energy sources, such as offshore wind turbines, are promising alternatives. However, the abnormal climate is a potential threat to the safety of offshore structures because construction guidelines cannot embrace climate outliers. A cosine similarity-based maintenance strategy may be a possible solution for managing and mitigating these risks. However, a study reporting its application to an actual field structure has not yet been reported. Thus, as an initial study, this study investigated whether the technique is applicable or whether it has limitations in the real field using an actual example, the Gageocho Ocean Research Station. Consequently, it was found that damage can only be detected correctly if the damage states are very similar to the comparison target database. Therefore, the high accuracy of natural frequencies, including environmental effects, should be ensured. Specifically, damage scenarios must be carefully designed, and an alternative is to devise more efficient techniques that can compensate for the present procedure.

Multiple Damage Detection of an Offshore Helideck through the Two-Step Artificial Neural Network Based on the Limited Mode Shape Data

A helideck is an essential structure in an offshore platform, and it is crucial to maintain its structural integrity and detect the occurrence of damage early. Because helidecks usually consist of complex lattice truss members, precise measurements are required for structural health monitoring based on accurate modal parameters. However, available sensors and data acquisition are limited. Therefore, we propose a two-step damage detection process using an artificial neural network. Based on the mode shape database collected from 137,400 damage scenarios by finite element analysis, the neural network in the first step was trained to estimate the mode shapes of the entire helideck model using the selected mode shape data obtained from the limited measuring points. Then, the neural network in the second step is consecutively trained to detect the location and amount of structural damage to individual parts. As a result, it is shown that the proposed procedure provides the damage detection capability with only a quarter of the entire mode shape data, while the estimation accuracy is sufficiently high compared to the single network directly trained using all mode shape data. It was also found that, compared to the network directly trained from the same data, the proposed technique tends to detect minor damages more accurately.

Fusing imperfect experimental data for risk assessment of musculoskeletal disorders in construction using canonical polyadic decomposition

Field or laboratory data collected for work-related musculoskeletal disorder (WMSD) risk assessment in construction often becomes unreliable as a large amount of data go missing due to technology-induced errors, instrument failures or sometimes at random. Missing data can adversely affect the assessment conclusions. This study proposes a method that applies Canonical Polyadic Decomposition (CPD) tensor decomposition to fuse multiple sparse risk-related datasets and fill in missing data by leveraging the correlation among multiple risk indicators within those datasets. Two knee WMSD risk-related datasets-3D knee rotation (kinematics) and electromyography (EMG) of five knee postural muscles-collected from previous studies were used for the validation and demonstration of the proposed method. The analysis results revealed that for a large portion of missing values (40%), the proposed method can generate a fused dataset that provides reliable risk assessment results highly consistent (70%-87%) with those obtained from the original experimental datasets. This signified the usefulness of the proposed method for use in WMSD risk assessment studies when data collection is affected by a significant amount of missing data, which will facilitate reliable assessment of WMSD risks among construction workers. In the future, findings of this study will be implemented to explore whether, and to what extent, the fused dataset outperforms the datasets with missing values by comparing consistencies of the risk assessment results obtained from these datasets for further investigation of the fusion performance.

Structural Monitoring of Underground Structures in Multi-Layer Media by Dynamic Methods

The actual problem of structural monitoring and modeling of dynamic response from buried building is considered in the framework of arbitrary dynamic load. The results can be used for designing underground transport constructions, crossings, buried reservoirs and foundations. In existing methods, the system of sensors that register the response to a dynamic action does not allow for effective interpretation of the signal without understanding the dynamic features and resonance phenomena. The analytical and numerical solution of the problem of the dynamics of a buried object in a layered medium is considered. A multilayer half-space is a set of rigidly interconnected layers characterized by elastic properties. At a distance, an arbitrary dynamic load acts on the half-space, which causes oscillations in the embedded structure, and the sensor system registers the response. The problem of assessing the dynamic stress-strain state (DSSS) is solved using Fourier transforms with the principle of limiting absorption. As an example, the behavior of an embedded massive structure of an underground pedestrian crossing under the influence of a dynamic surface source on a multilayer medium is considered, as well as instrumental support of the sensor system. The solution in the form of stress, strain and displacement fields is obtained and compared with the experimental data. The frequency-dependent characteristics of the system are determined and the possibility of determining the DSSS by a shock pulse is shown.

Structural Health Monitoring for Jacket-Type Offshore Wind Turbines: Experimental Proof of Concept

Structural health monitoring for offshore wind turbines is imperative. Offshore wind energy is progressively attained at greater water depths, beyond 30 m, where jacket foundations are presently the best solution to cope with the harsh environment (extreme sites with poor soil conditions). Structural integrity is of key importance in these underwater structures. In this work, a methodology for the diagnosis of structural damage in jacket-type foundations is stated. The method is based on the criterion that any damage or structural change produces variations in the vibrational response of the structure. Most studies in this area are, primarily, focused on the case of measurable input excitation and vibration response signals. Nevertheless, in this paper it is assumed that the only available excitation, the wind, is not measurable. Therefore, using vibration-response-only accelerometer information, a data-driven approach is developed following the next steps: (i) the wind is simulated as a Gaussian white noise and the accelerometer data are collected; (ii) the data are pre-processed using group-reshape and column-scaling; (iii) principal component analysis is used for both linear dimensionality reduction and feature extraction; finally, (iv) two different machine-learning algorithms, k nearest neighbor (k-NN) and quadratic-kernel support vector machine (SVM), are tested as classifiers. The overall accuracy is estimated by 5-fold cross-validation. The proposed approach is experimentally validated in a laboratory small-scale structure. The results manifest the reliability of the stated fault diagnosis method being the best performance given by the SVM classifier.

Structural Health Monitoring with Sensor Data and Cosine Similarity for Multi-Damages

There is a large risk of damage, triggered by harsh ocean environments, associated with offshore structures, so structural health monitoring plays an important role in preventing the occurrence of critical and global structural failure from such damage. However, obstacles, such as applicability in the field and increasing calculation costs with increasing structural complexity, remain for real-time structure monitoring offshore. Therefore, this study proposes the comparison of cosine similarity with sensor data to overcome such challenges. As the comparison target, this method uses the rate of changes of natural frequencies before and after the occurrence of various damage scenarios, including not only single but multiple damages, which are organized by the experiment technique design. The comparison method alerts to the occurrence of damage using a normalized warning index, which enables workers to manage the risk of damage. By comparison, moreover, the case most similar with the current status is directly figured out without any additional analysis between monitoring and damage identification, which renders the damage identification process simpler. Plus, the averaged rate of errors in detection is suggested to evaluate the damage level more precisely, if needed. Therefore, this method contributes to the application of real-time structural health monitoring for offshore structures by providing an approach to improve the usability of the proposed technique.