Structural Health Condition Research Articles

Framework for Identification and Prediction of Corrosion Degradation in a Steel Column through Machine Learning and Bayesian Updating

In recent years, structural health monitoring, starting from accelerometric data, is a method which has become widely adopted. Among the available techniques, machine learning is one of the most innovative and promising, supported by the continuously increasing computational capacity of current computers. The present work investigates the potential benefits of a framework based on supervised learning suitable for quantifying the corroded thickness of a structural system, herein uniformly applied to a reference steel column. The envisaged framework follows a hybrid approach where the training data are generated from a parametric and stochastic finite element model. The learning activity is performed by a support vector machine with Bayesian optimization of the hyperparameters, in which a penalty matrix is introduced to minimize the probability of missed alarms. Then, the estimated structural health conditions are used to update an exponential degradation model with random coefficients suitable for providing a prediction of the remaining useful life of the simulated corroded column. The results obtained show the potentiality of the proposed framework and its possible future extension for different types of damage and structural types.

Damage Estimation of Full-Scale Infilled RC Frames under Pseudo-Dynamic Excitation by Means of Output-Only Modal Identification

To assess the health condition of structures and infrastructure during their service lives, continuous vibration-based monitoring represents a viable and cost-effective solution. Model updating and digital twins are increasingly adopted for damage detection. However, significant gaps and uncertainties in damage quantification still arise. This work presents original data from output-only modal identification tests on full-scale, two-storey reinforced concrete (RC) frames subjected to pseudo-dynamic loading to simulate seismic damage. The frames are tested with two masonry infill wall configurations with three-sided and four-sided boundary conditions, and the observed seismic damage is correlated to a damage scale. Output-only modal identification tests are performed before and after testing to catch variations in modal properties due to observed damage. Experimental data are used to build a refined finite element model able to reliably simulate the static and dynamic performance of the infilled RC frames before and after damage. The model allowed for the further assessment of the variation in natural frequencies of tested specimens at different earthquake intensities, the correlation of such variations to damage levels, and identification of the contribution of structural and non-structural components to the overall frequency variation.

An Optimal Strain Gauge Layout Design for the Measurement of Truss Structures

Sensor measurements diagnose and evaluate the structural health state. A sensor configuration with a limited number of sensors must be designed to monitor sufficient information about the structural health state. The diagnosis of a truss structure composed of axial members can begin with a measurement by the strain gauges attached to the truss members or by the accelerometers and displacement sensors at the nodes. This study considered the layout design of the displacement sensors at the nodes for the truss structure by using the effective independence (EI) method based on the mode shapes. The validity of the optimal sensor placement (OSP) methods depending on their synthesis with the Guyan method was investigated by the mode shape's data expansion. The Guyan reduction technique rarely affected the final sensor design. A modified EI algorithm based on the strain mode shape of the truss members was presented. A numerical example was analyzed, showing that the sensor placements were affected depending on the displacement sensors and strain gauges. Numerical examples illustrated that the strain-based EI method without the Guyan reduction method has the advantage of reducing the number of sensors and providing more data related with the displacements at the nodes. The measurement sensor should be selected when considering structural behavior, as it is crucial.

Engineering surveys of Sri Lankan schools exposed to tsunami

The 2004 Indian Ocean tsunami affected 5% of Sri Lanka’s schools, severely damaging 108 and destroying 74. The catastrophe highlighted the critical role of schools in providing educational continuity during community recovery. Sri Lanka has since rehabilitated and rebuilt most of the destroyed schools along the coastline. However, there is a limited understanding of current levels of school exposure to tsunami. This hampers preparedness and risk reduction interventions that can improve community and educational tsunami resilience. This paper presents a multi-disciplinary school exposure database relevant to both vulnerability and loss modelling. The repository includes data on 38 schools and 86 classroom buildings, surveyed across the coastal districts of Ampara, Batticaloa and Galle in Sri Lanka, which were heavily affected by the 2004 tsunami. A new engineering rapid visual survey tool is presented that was used to conduct the physical assessment of schools for the exposure repository. School damage mechanisms observed in past tsunami inform the survey forms, which are designed to capture information at both school compound and building levels. The tsunami engineering survey tools are universally applicable for the visual assessment of schools exposed to tsunami. The surveys show that most Sri Lankan school buildings can be classified into three building archetypes. This means that future risk assessments can be conducted considering a small number of index buildings that are based on these archetypes with differing partition arrangements and structural health conditions. The surveys also raise three significant concerns. Firstly, most schools affected by the 2004 tsunami remain in the same exposed locations without any consideration for tsunami design or strengthening provisions. Secondly, Sri Lankan schools are fragile to tsunami loading and many of the schools in the Galle district suffer from severe corrosion, which will further affect their tsunami vulnerability. Thirdly, schools do not appear prepared for tsunami, and do not have adequate tsunami warnings nor evacuation protocols in place. These observations raise the urgent need to mitigate tsunami risk, including a holistic plan for tsunami retrofitting and for interventions to improve the tsunami preparedness of schools in Sri Lanka.

A State-of-the-Art Review of Non-Destructive Testing Image Fusion and Critical Insights on the Inspection of Aerospace Composites towards Sustainable Maintenance Repair Operations

Non-destructive testing (NDT) of aerospace structures has gained significant interest, given its non-destructive and economic inspection nature enabling future sustainable aerospace maintenance repair operations (MROs). NDT has been applied to many different domains, and there is a number of such methods having their individual sensor technology characteristics, working principles, pros and cons. Increasingly, NDT approaches have been investigated alongside the use of data fusion with the aim of combining sensing information for improved inspection performance and more informative structural health condition outcomes for the relevant structure. Within this context, image fusion has been a particular focus. This review paper aims to provide a comprehensive survey of the recent progress and development trends in NDT-based image fusion. A particular aspect included in this work is providing critical insights on the reliable inspection of aerospace composites, given the weight-saving potential and superior mechanical properties of composites for use in aerospace structures and support for airworthiness. As the integration of NDT approaches for composite materials is rather limited in the current literature, some examples from non-composite materials are also presented as a means of providing insights into the fusion potential.

Digital twining of an offshore wind turbine on a monopile using reduced-order modelling approach

In the wind energy industry, a cost-benefit digital twin (DT) will be very useful for managing the operation of a wind turbine. For instance, a DT could provide information on structural health conditions in real-time as well as projections for the near future. In this work, we employ a component-based Reduced-Order Modelling (ROM) technique to construct a DT of a parameter-varying offshore wind turbine system on a monopile. The DT consists of a ROM model for modal analysis and structural response prediction under wind and wave loadings. First, an offline library is pre-computed, which contains a series of component archetypes. Then, the component-based ROM is assembled out of these archetypes, such as the parts in the blade, hub, nacelle, and tower. With the computation speed of two orders (approximately 650x) faster than a Finite Element Analysis (FEA) model and high accuracy (less than 0.2% error), the DT could be able to provide almost instant predictions of the modes and the responses of the turbine structure due to the wind-wave loading as well as projections of structural health conditions.

A Systematic Review of Advanced Sensor Technologies for Non-Destructive Testing and Structural Health Monitoring.

This paper reviews recent advances in sensor technologies for non-destructive testing (NDT) and structural health monitoring (SHM) of civil structures. The article is motivated by the rapid developments in sensor technologies and data analytics leading to ever-advancing systems for assessing and monitoring structures. Conventional and advanced sensor technologies are systematically reviewed and evaluated in the context of providing input parameters for NDT and SHM systems and for their suitability to determine the health state of structures. The presented sensing technologies and monitoring systems are selected based on their capabilities, reliability, maturity, affordability, popularity, ease of use, resilience, and innovation. A significant focus is placed on evaluating the selected technologies and associated data analytics, highlighting limitations, advantages, and disadvantages. The paper presents sensing techniques such as fiber optics, laser vibrometry, acoustic emission, ultrasonics, thermography, drones, microelectromechanical systems (MEMS), magnetostrictive sensors, and next-generation technologies.

A numerical study on planar gradient acoustic impedance matching for guided ultrasonic wave detection

Increasing the service-life of engineering structures such as aeroplanes is a major issue in order to enhance their cost-effectiveness and to reduce the carbon footprint. One possibility to achieve this goal is to determine the current structural health state and to derive respective measures in order to increase the structure’s technical reliability. For doing so, a structural health monitoring system consisting of both an actuator and a sensor network may be applied. Whereas the actuator induces a wave field of guided ultrasonic waves, the measuring data of the sensors allows to determine the health state of the respective structure. However, both actuators and sensors in most cases distort these wave fields. This distortion may lead to false-detection of damage: both the number and severity of damage may be over- or underestimated. The former leads to an unnecessary high effort for retrofitting the structure, whereas the latter reduces the structure’s technical reliability. Several measures exist in order to avoid such false-detections. In the present contribution, focus is set on reducing the distortion of the wave field which is caused by an embedded sensor. The reduced distortion of the wave field is achieved by an acoustic impedance matching with a functionally graded material which is based on a mechanical model. The approach additionally results in amplified measuring signals of the sensor. The applicability of the proposed approach is shown by means of a numerical study.

Dynamic identification of an elevated water tank through digital video processing

Ageing structures and infrastructures need to be monitored to assess their structural health conditions and prioritize interventions to possibly extend their service life. To do this, accelerometers and velocimeters are routinely adopted and are part of a consolidated state-of-the-art procedure. Nonetheless, some difficulties may arise in field applications, related to energy supply, cost, and accessibility of the devices. Moreover, the position of the sensors needs to be decided a priori, with some degree of engineering judgment. Alternative techniques based on computer vision have emerged in the last decade and are becoming more and more popular as they allow to overcome most of the limitations reported above. The main advantages of these approaches rely on the possibility of high-density measurements and a relatively simple acquisition process, for which neither an expensive equipment nor advanced technical skills are mandatorily required. In this paper, a computer vision-based technique is presented, which combines motion magnification and statistical algorithms. It was applied to extract the natural frequencies of a reinforced concrete elevated water tank, vibrating under environmental noise excitation. To this aim, several videos were recorded with a commercial reflex camera and post-processed selecting a representative area, by tracking in time either the variation of the intensity or the motion of a selected number of pixels. Computer vision-based outcomes were validated against the results provided by accelerometers to discuss advantages and limitations of the proposed dynamic identification approach and identify future research challenges in this field.

A Damage Detection Method Based on Flexible Macro-Fiber Composite

Macro-fiber composite (MFC) transducers exhibit superior directionality and flexibility, as well as a high response to Lamb waves, which endows them an intrinsic advantage in damage detection. This article applies MFC transducers to “active–passive” damage detection, a technique which is not well-documented, on a curved structure. By rationally laying out the actuators and rosette receivers on a curved structure of interest, we are able to detect the baseline-free location of infinitesimal damage in the structure. This study combines computational simulations, theoretical analyses, and experimental validations to investigate directional excitation and reception characteristics, in which the incoming wave direction estimation algorithm of MFC delta rosette structure is obtained and damage is characterized by the intensity image synthesized from error estimation curves of two MFC rosette receivers. Experimental results show that the location errors in pinpointing the damage positions on the curved aluminum plate are all within 10 mm, and this system can effectively detect damage in the structure with up to 5 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\text{m}^{-1}$ </tex-math></inline-formula> curvature. Our preliminary results demonstrate that MFC has the potential to precisely locate damage on complex curved structures and provides a novel testing technique to monitor the health state of complex structures.

An Improved Structural Health Monitoring Method Utilizing Sparse Representation for Acoustic Emission Signals in Rails

The structural health of rails significantly impacts the safety of railway transport, which requires long-term and accurate monitoring. An improved structural health monitoring (SHM) method is proposed to enhance the damage information and construct an adaptive weighted index for evaluating the structural health states of rails. In this method, modified singular value decomposition (SVD) with a double-layer Hankel matrix compensation algorithm enhances the damage information and ensures real-time detection. Based on the inflection point of the inner product gradient, the step size of the sparsity adaptive matching pursuit (SAMP) algorithm is adaptively adjusted to reconstruct damage signals. Subsequently, an innovative index is constructed with adaptive weighting for SHM of rails. Tensile tests are performed to verify the effectiveness of the proposed method. The results show that the proposed method avoids the defects of traditional nondestructive testing (NDT) techniques that rely on a priori information about the damage signal (such as damage frequency acoustic emission (AE) hits rate and waveform characteristics). Furthermore, using the variation in signal energy, the proposed method has the capability to accurately delineate the three structural health stages under complex waveform conditions, which is not possible with conventional detection methods. This method provides guidance for the damage analysis and monitoring techniques of rails.

Seismic damage and loss evaluation in precast industrial buildings through low-cost accelerometers

Past seismic events, in particular the Emilia earthquake (2012), showed how the damage caused in highly industrialized areas could cause serious problems with large-scale economic effects. It is interesting to highlight how these effects are both immediate, due to the seismic vulnerability of old precast industrial buildings, and in the long term since the restoration operations are typically long due to the difficulty of assessing the real structural health state, particularly for the difficulty in reaching and inspecting components sensitive to damage, such as the connections. This work has the objective of finding a procedure, through relatively low-cost MEMS accelerometers, which allows to record and process the seismic accelerations experienced by the various structural elements. The paper considers the application of a procedure that allows identifying the state of health of the system through the application of damage indexes found in the literature and a first estimate of costs and recovery times to bring the building back to its pre-earthquake conditions. The procedure has been applied to a selected case study

Satellite interferometric data for seismic damage assessment

Radar satellites allow the collection of data on large areas without direct access to structures. Thereby, they appear very attractive for Structural Health Monitoring (SHM) purposes. Data collected by satellites can be processed to obtain temporal histories of displacements through which the health state of a monitored system can be potentially identified. However, anomalies in the time histories of displacements are not necessarily due to damage. Environmental phenomena, such as variations in atmospheric temperature, and rain, can modify the behavior of structures without compromising their safety. The impact of these phenomena on the structural response can hinder the identification of anomalies or lead to false alarms if such alterations are misinterpreted as damage. Furthermore, if the monitored system is a historical structure, uncertainties on the structural behavior are inevitably increased during aging. The purpose of this article is to discuss the possibility of identifying damage due to seismic actions considering the impact of variations of environmental factors on the time histories of the displacements retrieved by satellite data. The structural health condition of a historical structure located in the city of Rome (Italy) hit by the October 2016 Central Italy earthquakes is investigated based on interferometric satellite data. The satellite data are acquired by COSMO-SkyMed (CSM) of the Italian Space Agency between 2010 and 2019 and are processed by CNR IREA.

Multi-Vib

Vibration measurement is vital for fault diagnosis of structures (e.g., machines and civil structures). Different structure components undergo distinct vibration patterns, which jointly determine the structure's health condition, thus demanding simultaneous multi-point vibration monitoring. Existing solutions deploy multiple accelerometers along with their power supplies or laser vibrometers on the monitored object to measure multi-point vibration, which is inconvenient and costly. Cameras provide a less expensive solution while heavily relying on good lighting conditions. To overcome these limitations, we propose a cost-effective and passive system, called Multi-Vib, for precise multi-point vibration monitoring. Multi-Vib is implemented using a single mmWave radar to remotely and separately sense the vibration displacement of multiple points via signal reflection. However, simultaneously detecting and monitoring multiple points on a single object is a daunting task. This is because most radar signals are scattered away from vibration points due to their tilted locations and shapes by nature, causing an extremely weak reflected signal to the radar. To solve this issue, we dedicatedly design a physical marker placed on the target point, which can force the direction of the reflected signal towards the radar and significantly increase the reflected signal strength. Another practical issue is that the reflected signal from each point endures interferences and noises from the surroundings. Thus, we develop a series of effective signal processing methods to denoise the signal for accurate vibration frequency and displacement estimation. Extensive experimental results show that the average errors in multi-point vibration frequency and displacement estimation are around 0.16Hz and 14μm, respectively.

Real-time drive-by bridge damage detection using deep auto-encoder

Structural health condition monitoring of bridge structures has been a concern in the last decades due to their aging and deterioration, in which the core task is damage detection. Recently, the drive-by method has gained much attention as it only needs several sensors installed on the passing vehicle. In this paper, we proposed an automatic damage detection method, which can be exploited in real time when the vehicle is passing the bridge. There are three steps in the proposed method: (1) The vehicle’s framed short-time vibrations instead of full-length data are utilized for training a deep auto-encoder model; at this stage, not commonly used time-domain accelerations of the passing vehicle, but its selected frequency-domain responses are employed to circumvent the influence of noises, (2) For the bridge with unknown health conditions, damage indicators can be extracted from its passing vehicle’s short-time vibration data using the trained model, and (3) The bridge’s health states are determined by real-time extracted damage indicators. To verify the proposed idea, a U-shaped continuous beam and a model truck are used to simulate the vehicle bridge interaction system in engineering. Results showed that the proposed method could identify the bridge’s damage with an accuracy of 86.2% when different severity was considered. In addition, it was observed that higher damage severity could not be revealed by greater values of damage indicators in the laboratory test. Instead, a novel index called identified damage ratios was employed as a reference for assessing the severity of the bridge’s damage. It was shown that with the increase in damage severity, the index would increase and gradually approach 100%.

Linear annular path damage probability distribution based ultrasonic guided wave method for position imaging and tracking of multi-damage on plate-like carbon fiber composite structure

Plate-like carbon fiber composite is an important component of the main bearing structure of spacecraft and high-speed trains, which has significant advantages in reducing the dead weight and energy consumption, and improving the carrying capacity. However, due to long-term exposure to complex service environments, material properties will be changed, these changes accumulated to a certain extent will lead to structural damage, threatening equipment operation and personal safety. To facilitate the attainment of mastering structural health states, this paper proposes a damage position imaging and tracking technology based on linear annular path damage probability distribution and ultrasonic guided wave for composite plate-like structures. Firstly, according to the dispersion curve, the propagation velocity characteristics of guided wave on composite structure are analyzed. Then, combined with the limited number of measured propagation velocities, the group velocity of the A0 mode in each direction is fitted. Based on the fitted velocity, the time-of-flight (TOF) map for all search points on each path can be pre-constructed. Subsequently, the damage TOFs and absolute TOF difference threshold are combined to obtain the annular damage probability distribution map of each sensing path. Ultimately, the total damage probability of the whole monitoring region is obtained by summing up the path damage probability, which can directly reflect the current damage position. Experimental results show that the proposed method has obvious advantages in imaging and tracking of single or multiple damage positions. It is hoped that the findings in this paper can provide new ideas for long-term damage state monitoring of the composite structures.

Condition Assessment and Adaptation of Bailey Bridges as a Permanent Structures

The present study assessed the Bailey Bridge’s condition and investigated its adaptation as a permanent structure, targeted the Acrow Bailey Bridge in Japan. Field diagnostic loading experiments were performed under various loading conditions, such as dynamic and static loading tests. The onsite data were obtained using a transducer, friction strain gauge, target measurements for the image processing approach, and accelerometer. From the field measurements, the deflection and stresses of the bridge were found to operate within the linear elastic region. The bridge was then accurately modeled based on the in situ geometric configuration of the bridge, and Finite Element Analysis was performed. The model’s accuracy was validated with the onsite data under the linear elastic domain. The model was deployed to check for resistance of critical members. A nonlinear analysis based on the linear and nonlinear buckling method was performed to determine the subject bridge’s Serviceability Limit State and Ultimate Limit State. The results showed that the first out-of-plane eigenvalue buckling analysis could monolithically assess bridge members. Further, the study established digital twin models resolve for historical data through in situ modeling measurements. Therefore, the findings obtained in this study highlight the bridge’s Structural Health Condition, bearing capacity, and propose a framework for adaptation as a permanent structure.

A novel deep convolutional image-denoiser network for structural vibration signal denoising

Vibration-based approach is of great importance for structural health monitoring and condition assessment, while inevitable noise existing in field measurement casts great obstacles in corresponding data-driven analysis. It has been a stringent prerequisite to develop effective methods to denoise vibration signal. Hence, a novel denoising approach based on deep convolutional image-denoiser networks (DCIMN) is proposed in this study, the methodology and architecture of which are elaborated. Specified avenues with novelties including noise injection in training labeling, dimension expansion in feature extraction, and optimizer embedding in encoder–decoder are utilized to enhance the denoising performance. Measured vibration data from Shanghai Tower is allocated for validation, based on which modal identifications are also conducted. Detailed evaluation confirms its powerful capability and efficiency in denoising signal. Demanding no prior information of input signal, the proposed method performs vibration signal denoising in an intelligent way, which demonstrates a vast prospect in engineering practice.

Strain based structural health monitoring of stiffened composite plate

Stiffened composite structures are very appealing in aeronautical applications due to their unique stiffness to mass ratio. However, they are prone to various and complex damage scenario such as stiffener debonding, impact damages etc. The composites materials suffer from hidden damages that lead to catastrophic failure if not detected at an early stage. This new breed of structures (advanced composites structures), therefore pose more challenges in predicting their remaining life or state of structural health. Consequently, autonomous monitoring of such structure is still a real issue. The current study develops a structural health monitoring approach that uses sensors with low-cost strain gauges to identify damage. A parametric finite element model is used to generate multiple damage scenarios for the investigation. To determine the exact placement of strain gauges, a numerical analysis is performed and longitudinal strain values are obtained in both healthy and damaged states. The experimental testing of the stiffened composite plate is carried out on healthy and damaged condition. Results from numerical simulations have been compared with experimental result. The result of this work indicates that the strain gauge can be used as effective tool to detect the damage from their early development, location and severity of damage that makes it particularly useful for structural health monitoring.

LSTM approach for condition assessment of suspension bridges based on time-series deflection and temperature data

Deflection data provides important information about the mechanical characteristics and structural health condition of bridges. The study presented here pertains to development of a deep learning based approach for structural health monitoring by employing the bridge deflections. The method presented herein uses the long short-term memory (LSTM) framework in detecting the state of damage by tracking the feature changes of time-series deflection and temperature data. Deflection and temperature data of Chongqing Egongyan Rail Transit Suspension Bridge was employed over a period of 15 months to develop the proposed method. The concept of square error index (SE) is introduced as an assessment tool for estimation of the bridge damage level. Results from the present study indicated that the statistical characteristics of SE index are proportional to the level of damage, and are only sensitive to abnormal changes in deflection. Structural health monitoring data over the period of 15 months indicated that the proposed approach has the capability to detect cable damages as low as 0.5%.