European Workshop on Structural Health Monitoring

Recently, Instantaneous Spectral Entropy (ISE) measurements have been proposed for real-time damage assessment purposes in nonstationary mechanical systems. These include Condition Monitoring (CM) in rotating machinery as well as Structural Health Monitoring (SHM) for civil structures. However, several distinct entropy definitions are available in the scientific literature, each with its advantages and limitations. In the present paper, the potential of the family of Rényi entropies is evaluated. The order $$\alpha$$ α is varied between 0 (corresponding to the Hartley entropy, i.e., the Max-entropy) and $$+ \infty$$ + ∞ (Min-entropy), encompassing $$\alpha \to 1$$ α → 1 (Shannon-entropy) and $$\alpha = 2$$ α = 2 (Collision entropy) as particular cases of special interest. The Wigner-Ville (WV), Smoothed Pseudo Wigner Ville (SPWV), Discrete Choi-Williams (DCW), and Continuous Wavelet (CWT) transforms are all tested for the time-frequency (TF) distribution of the target signal. In turn, this TF is used to compute the probability density function, from which the ISE is estimated. A sensitivity analysis is run on all the parameters for each candidate TF transform, aiming at defining the best settings. All studies were performed on a well-known experimental dataset, the three-storey aluminium frame structure developed at the Los Alamos National Laboratories, considering the undamaged conditions as the normality model. The results show good potential for entropy-based real-time damage detection, especially for breathing cracks.

The use of wavelets is a popular tool in vibration-based damage detection due to their ability to identify the changes in the instantaneous frequency of structures. Despite their widespread use in damage detection, few studies systematically investigate the efficacy of different waveforms in identifying variations in instantaneous frequency and damage detection. This article aims to address this gap in the literature. Wavelet properties, like center frequency and vanishing moment, are of importance for the wavelet selection process. For this, first, the moving load analysis was performed on the numerical model of the undamaged and damaged bridge with a low level of damage, and acceleration time histories were obtained at different observation points. The continuous wavelet transform (CWT) using different wavelet functions including Symlets, Daubechies, Coiflets, Gaussian, were applied on the acceleration time histories. Then, the absolute difference of wavelet coefficients between damaged and undamaged acceleration signals were used in order to evaluate the success of each wavelet in damage detection. At the end, different resolutions were observed in high and low scales by using different vanishing moment, and by use of a wavelet coefficient line plot, Gaus2 showed the more sensitivity to detect the damage location compared to the others.

Due to planned renovations for an important Broadcasting studio in Vienna, Austria, vibration monitoring was proposed in the adjacent technical server rooms due to the particularly sensitive requirements. Since the importance played by this Corporation as one of the largest media provider in the country, it must be continuously operational and must respect the latest vibration standards. To assess ambient vibrations, a first measurement was carried out on February 18, 2019 in the affected areas of the building. This included an assessment of the ambient and random vibrations. Moreover, a short-term measurement was carried out during test construction works to analyze the excited frequencies in the server rooms.Thus, relevant vibration events were measured and evaluated. The determined impacts were compared with the protection goals from relevant national and international standards and guidelines to give a recommendation on limiting the impacts for the future construction phase. Based on the analysis of various norms, a proper set of limits was developed, tailored to this case, and was finally applied. Then, a permanent monitoring system was installed with a set of alarms in real-time to warn and take measures in case thresholds are exceeded. Results are here also presented.

The Éric Tabarly bridge in Nantes is a 210 m long cable-stayed road bridge crossing the Loire River, that was inaugurated in 2011. It is composed of a 27 m wide steel deck divided into two spans by a 57 m high steel pylon, being the main span 143 m long. In the context of the European Research Project DESDEMONA (DEtection of Steel DEfects by Enhanced MONitoring and Automated procedure for self-inspection and maintenance), the bridge has been equipped with a dynamic monitoring system, constituted of 16 uniaxial accelerometers installed both on the deck and on the Pylon, with accelerations being recorded with a sampling rate of 100 Hz. This paper describes the dynamic monitoring system installed in the bridge and the results achieved during the first months of operation, including the characterization of vibration levels (maximum and effective values) as well as the automatic identification of the bridge modal properties. Finally, the effects of operational and environmental conditions on modal properties are studied.

A fiber-optic based structural monitoring system has been designed and deployed by CETENA of the Fincantieri Group on the new Genoa San Giorgio bridge, which replaced the collapsed Polcevera viaduct. A network of more than 240 sensors, most of them based on fiber-optic technology, has been installed to follow movements, vibrations and operating conditions of the bridge deck and piles in real-time. First of all, the monitoring solution aims to keep under control the loads and deformations of the structure, for verifying that its behavior lies within the boundaries of the project design. Both the bridge displacement and strain (due to the temperature changes and the passage of vehicles) and the piles’ tilt are recorded and analyzed to relate the actual behavior of the structure. In addition, several accelerometers are distributed along the bridge deck to check the bridge vibrations with Operational Modal Analysis (OMA). OMA allows to reconstruct the real structural dynamic behavior in real-time and compare it to the project design one, depending on the boundary conditions. At the same time, a database on both the structure behavior and the surrounding conditions is being built for properly and efficiently planning maintenance activities in the medium and long term. In summary, the fiber-optic structural monitoring system is the solution chosen by the owner for both efficiently operate the infrastructure and assuring the safety of people over many decades.

Large infrastructure projects can have an impact on the structural health of existing buildings. For example, the boring of tunnels can affect infrastructures located on the surface. Therefore, devices are required for the monitoring of every building in the area of interest.Topographic surveying is one of the most popular techniques for the monitoring of buildings and civil infrastructures. Standard optical prisms are targeted with high frequencies by Robotic Total Stations (RTS). In order to estimate 3D displacements of each prism, static control points are needed. These points must be located outside the area of interest. However, it is not always possible in dense urban areas.In this study, the use of GNSS receivers as control points for topographic surveying is proposed. In this case, control points do not need to be static because their global positions are known. The methodology is explained and the required additional processing steps are outlined. The approach is implemented for the monitoring of a conveyor belt in the Grand Paris Express (GPE) project in Paris, France. Feedback is provided on implementation, data collection and processing. Finally, performance of the system is analyzed and compared to traditional monitoring techniques.

Due to growing traffic demand, aging civil infrastructure raises the need for reliable tools to monitor structural health conditions, usable to plan informed maintenance and emergency management. Several structures with historical and monumental importance are instrumented with structural health monitoring (SHM) systems nowadays. However, even the failure of “minor” viaducts could endanger the safety of travelers and goods. Lately, dense wireless sensor networks (WSNs) based on MEMS devices are used to cut costs and simplify the deployment of SHM systems while collecting as much information as possible. However, dense WNSs are affected by data management, synchronization, and battery replacement issues, which make them unappealing for widespread use. This study presents an original damage identification algorithm based on sparse sensor networks. Traveling vehicles are exploited to obtain spatial information and accurately identify the location of structural anomalies. The curvature influence line of the monitored bridge can be calculated by processing the acceleration response measured at a given instrumented location through a low-pass filter. In this procedure, sensors operate individually, not needing energy-consuming synchronization. The proposed identification algorithm is verified on real data collected on a steel truss bridge subject to artificially induced damage.

Road bridges are facing the effects of both their ageing and the increase of the traffic loads, which justifies the necessity of combining Weigh-In-Motion (WIM) with Structural Health Monitoring (SHM) solutions. An operational case study located in southern Italy is described, on which a single Monitoring System achieves both results: to perform Bridge-WIM by using the bridge deck as a scale for the detection and qualification of the traffic load, and to assess the structural health of the asset through combined strain and vibration analysis. The system uses a low number of sensors and is not intrusive for the pavement, making it an efficient and comprehensive solution for the asset manager. The Salso viaduct is a concrete structure with nine simple-supported 32 m long spans, composed of four main prestressed concrete beam girders and a reinforced concrete slab. The system has been installed on two spans with twelve sensors, among them eight optical strand strain gauges and four accelerometers. Two cameras identify the weighed vehicles. Moreover, each truck crossing the bridge triggers strain and vibration records and these data are further analyzed for the SHM of the concrete structure. The performance of the Bridge-WIM system was assessed according to the COST323 specification and the OIML R-134 recommendation. It reached the COST323 accuracy class A(5) and the OIML class 10 for the gross vehicle weights. The OIML class 5 was reached for the fully loaded trucks.

The Civil Engineering Building was built between 2017 and 2019 to provide new office and laboratory space for the Engineering Department at the University of Cambridge, and to house the National Research Facility for Infrastructure Sensing (NRFIS). During the construction, monitoring systems using distributed fibre optic sensing (DFOS) and fibre Bragg grating (FBG) technologies were installed in four different parts of the building: ground source heat pump boreholes, the basement raft slab and retaining walls, the structures lab strong floor slab, and the steel columns and beams of one of the building’s load-bearing frames. Software packages to acquire, store, analyse and visualise data in real time were also developed as part of the monitoring systems. The monitoring systems are described in this paper, along with some of the initial data recorded during and soon after construction.

As space exploration programs around the world continue with accelerated plans for planetary robotic missions and human expeditions to Mars, the Moon, and beyond, laying the groundwork for even more complex human science expeditions, the need for spacecraft to land safely on planetary surfaces has become increasingly challenging because of the use of massive and hauling larger payloads required to accommodate the extended stays on the Martian and Lunar surface. Advances in supersonic decelerator technology investigates re-entry vehicle designs that evaluate reliable techniques for safe planetary atmospheric re-entry. Decelerator design engineers are investigating the use atmospheric drag as a solution to save rocket engines and fuel for final manoeuvres and landing procedures. The heavier planetary landers of tomorrow will require much larger drag devices use to slow them down during re-entry deployed at higher supersonic speeds to safely land vehicle, crew, and cargo. Aerial entry, descend, and landings system model validation and verification are an area in the aerodynamic decelerator community that is constantly growing and innovating.

Delamination is a serious problem in concrete bridge decks. Impact sounding has proved to be one of the most effective techniques for defect detection on concrete surfaces. However, the data collection technique is generally compromised due to the noise associated either with the surroundings or the equipment itself. The data collection is also affected due to human errors. This paper studies different data collection techniques using Hammer (flat and round), Robotic crawler, and Laser Doppler Vibrometer (LDV) on a concrete slab. The slab is embedded with different types of artificially constructed defects, such as shallow delamination, void, honeycomb, and deep delamination. Empirical Mode Decomposition (EMD) has been applied to filter the noise from the data collected. Different advanced signal processing techniques such as Power Spectral Density (PSD), Hilbert Huang Transform (HHT), and Hilbert Marginal Spectrum (HMS) have been applied for defect detection, and the results have been compared. The limitations of the different data collection techniques and signal processing techniques have also been discussed.

Tainter gates are commonly used as water control gates on civil infrastructure, with one such gate in use at the lock and dam at The Dalles, Oregon on the Columbia River. Under normal operating conditions, Tainter gates should be hoisted in a level fashion throughout the circular path they travel. Uneven hoisting may lead to a redistribution of loads or accelerated fatigue damage, and an instrumentation system was installed on the subject gate with a goal to monitor for such hoisting. Additionally, a numerical model was created to investigate changes in structural response when the gate hoists unevenly. Comparison of the initial data to the model suggested that the gate was regularly hoisting unevenly. To increase confidence in the instrumentation system, and aid in validating the numerical model, short-term monitoring was leveraged wherein the hoisting cables of the gate were instrumented with accelerometers. Then, a frequency domain method was used to infer the tension in each of the cables that hoist the gate, which is expected to be approximately equal if the gate is hoisting evenly. Results from the monitoring program support the notion that the gate is regularly hoisting unevenly, thus increasing confidence in the instrumentation system.

Most modern seismic design codes build upon the concept of performance-based earthquake engineering that allows structures to sustain repairable damage during moderate and large earthquakes. Therefore, accurate and quantitative post-earthquake damage evaluation of real-world structures is crucial for safe operation of buildings. Structural-health monitoring provides sensor-based information regarding the structural state and informs post-earthquake building assessment. With the utilization of monitoring data, which is recorded during earthquake excitation, damage-sensitive features (DSFs) can be extracted in both purely data-driven or hybrid forms; with the latter term referring to damage indicators (DIs) that fuse data with dynamic models. In this paper, data-driven and hybrid damage identification methods are introduced and compared with respect to their performance and robustness in detecting and quantifying structural damage. The damage localization and quantification performance are discussed for varying number of building floors. Moreover, numerical models are used to enable the comparison of DSFs with metrics of nonlinearity, such as maximum drift, and with response metrics that are traditionally used to quantify damage, such as maximum inter-story drifts. Finally, uncertainties in DSFs and their sensitivity to sensor noise, prior knowledge of mass and the spectral content of earthquake excitation are assessed to explore the robustness of the hybrid DI.

In the context of climate change, environmental actions on structures are likely to alter in terms of intensities and frequencies of occurrence. To ensure sufficient load-bearing capacity of structures despite these changes, actions may be monitored using structural health monitoring (SHM) systems. Environmental actions involve time-dependent and non-scheduled loads, e.g., wind and snow loads. In current SHM systems, these loads are mostly traced locally. However, local monitoring may cause inaccuracies, as certain load phenomena, such as wind turbulences, or snow accumulations in specific parts of structures, may not be registered. A holistic, global recording of loads acting on structures has rarely been established since a multitude of sensors is cost intensive, and the integration into the building envelope is challenging. This paper investigates slender layered piezoresistive sensors to measure loads resulting from environmental actions, focusing on wind and snow loads. The sensors operate based on changes of externally applied pressure, leading to variations in the electrical resistance of a piezoresistive material. Next to strategies for quantifying structural loads using sensor technology, first, alternatives of force sensors are discussed. Subsequently, the low-cost technical fabrication of the piezoresistive pressure sensors is presented, and implementation, calibration, and validation of the pressure sensors are conducted. Finally, the validation results of the sensors are discussed, and an outlook on future work is presented. In summary, the sensors investigated offer a wide range of applications for monitoring structural actions on surfaces, serving as a basis for estimating the load-bearing capacity of structures reliably.

The paper reviews the monitoring, data analysis and modelling of a multi-span RC-bridge. The bridge is 102 m long, with six spans, crossing a glacial river in SE Iceland. The bridge was taken out of operation in 2019 when one of the bridge pillars subsided due to scour erosion. The Icelandic Road and Coastal Administration decided to use the bridge as a learning tool before it was demolished and a new bridge built on the same site. A realistic assessment of the structural properties and bearing capacity of existing bridges is a prerequisite for evaluating the safety of older bridges in the road system, which were designed according to outdated standards and different load conditions but are required to serve today’s traffic and future demands with an increasing number of heavy vehicles. The data presented in this paper are time series of acceleration measured for ambient loading on the bridge. A system identification analysis of the bridge was performed using two OMA tools. It is found that based on low amplitude ambient vibration data, the first four modes can be reliably estimated in terms of natural frequency and critical damping ratio, but the higher modes are more uncertain, and their identifiability varies depending on the data-set used. The results of the analysis are used to update and validate finite element models of the bridge that are used to interpret and simulate the observed vibrations.

Civil structures are designed to withstand loads from multiple sources; however, design is achieved based on several hypotheses, which in some cases do not occur. Phenomena such as differential settlements, damage or failure of structural elements, overloads, etc., cause buildings to behave differently than expected. When any of these phenomena occur, due to multiple factors, it is not possible to determine precisely the distribution of stresses in the building and therefore, it is difficult to establish with any degree of reliability if there are elements that may be subjected to stress levels that could cause a critical failure of the structure. In this paper, the development of a low-cost fiber optic-based transducer intended to force/stress measuring in civil structure pillars within the SHM context is presented. Test transducers were embedded in scale reduced columns, which were submitted to compression loads using a testing bench rig. Transducers successfully measured strain levels up to 200 $$\upmu \upepsilon $$ μ ϵ under compression load with an accuracy of ±1 $$\upmu \upepsilon $$ μ ϵ . Strains and applied force were linearly related. Through finite element analysis (FEA), a sensing scheme was designed for the first floor of a 32-story-tall building composed of 32 columns where each column includes one embedded transducer. The constructive phase of building will start by April 2022, when transducers will be embedded within the structure. It is expected to detect load redistribution during the construction phase of the building and its operational phase.

The incorporation of Long Welded Rail (LWR) or Continuous Welded Rail (CWR) on bridges has led to the importance of Rail Structure Interaction (RSI) Analysis. It largely deals with determining the forces caused in rails and bridge components due to contact effects and checks for the safety of track and structural elements. Three major cases considered for generating additional stresses in rails due to deck under the tracks are (i) thermal expansion or contraction of bridge decks or tracks, (ii) deflection of sub-structure under tractive or braking forces from the trains, and (iii) end rotations caused by vertical bending under vertical trainloads. In this study wireless sensor network system is installed on the first span of the bridge with LWR at various locations to measure the additional stresses obtained from three tests (i) sudden braking of the train on the bridge, (ii) deformation of the bridge for vertical moving load and (iii) 24-h temperature change. Results obtained were compared with the numerical model generated using finite element method-based MIDAS software. Variations in the results and challenges in the instrumentation of the bridge for acquiring the data are presented in detail.

The Brivio bridge, completed in 1917, consists of three reinforced concrete (RC) tied arches spanning 44.0 m each. The bridge crosses the Adda river on the route between Lecco and Bergamo and still represents a crucial node for the vehicular traffic of the region network. Hence, in the context of a research promoted by the Lombardy Region, the bridge has been equipped with a dynamic monitoring system, consisting of 8 seismometers per span. The response to operational excitations is collected at a sampling frequency of 100 Hz, with datasets of 3600 s being created every hour for automated processing.The paper describes the dynamic monitoring system installed in the bridge and selected results obtained during the first year of operation. Overall, seven vibration modes are identified for each span, with the temperature significantly affecting the natural frequencies, whereas no remarkable changes of mode shapes are detected so far. The possible onset of structural changes is identified by using both the changes in mode shapes and the cleansed natural frequencies.

Structural Health Monitoring (SHM) has been a significant research topic to help with damage detection in civil structures and to stop further deterioration. Traditional methods of SHM are time consuming and cost ineffective. In addition, civil structures such as dams and high raised buildings are burdensome and risky to inspect manually, especially after a natural disaster. Crack signals the beginning of failure for any structure. Most of the existing methods largely deal with only the detection of cracks. Proposed work concentrates on segmentation, classification, and subsequent detection of cracks based on pattern i.e., Linear vs branching, apart from the single and multiple cracks. The image dataset was obtained from real-time visual inspections. This study is significant because a branching crack shows greater structural stress than a linear crack. Furthermore, results quantify the damage in the image using instance segmentation techniques. Experimental analysis achieves classification and quantification of the data with good accuracy.

Palazzo Lombardia is one of the tallest buildings in Milan and the site of the regional government. For these reasons, some years ago a monitoring system was installed in order to assure its continuous operation. The system is based on accelerometers and clinometers at different floors used for dynamic and static monitoring, respectively.A statistical model was developed, such that the time trend of the first eigenfrequencies of the building were modelled through the trend of the clinometer signals and the root mean square (RMS) of some of the accelerometers. This because it was observed that the clinometer signals and the acceleration RMSs are linked to different environmental variables. As examples: the changes of the static configuration of the building due to sun exposure can be described by clinometer signals and acceleration RMSs can take into account the effect of wind. The use of these signals and indexes simplifies the development of the predictive model, compared to the use of measured environmental quantities. The model showed good performances in foreseeing the trend of the first eigenfrequencies.This paper analyses how the reliability of the model, developed with data acquired in 2015–2016, has changed relying on new data acquired in 2021–2022.

Structural Health Monitoring (SHM) has grown to improve human safety and reduce Infrastructural maintenance costs. It aims to detect, locate and quantify damage to the structure through data acquisition. However, most of the existing SHM systems face challenges performing in real-time due to environmental effects and different operational hazards. Therefore, the use of smart connected devices through the internet has provided flexibility to monitor live structures (buildings, bridge, etc.) from anywhere. These Internet of Things (IoT) devices can be linked to cloud platforms such as AWS for monitoring and analyzing the data. In this paper, a complete architecture is described to enable live monitoring applications on the AWS cloud platform. However, IoT is the network of physical devices that can send and receive data wirelessly without human intervention. IoT allows objects to be sensed and controlled remotely across the existing network infrastructure. Sensors (Accelerometers) use the Xbee protocol to transfer data to a common node (raspberry pi), then to the cloud using messaging protocols. The objective of the work demonstrates the architecture to carry out data collection, analysis, and visualization of sensor data using AWS services that result in the development of live monitoring standalone web interface.

In recent years, interest has grown in the maritime sector towards transitioning from preventive to predictive maintenance procedures. In terms of hull structure maintenance, the incorporation of Structural Health Monitoring techniques has emerged as a leading option. The objective of this work is to move towards this direction by investigating the use of classical detection theory tools to detect corrosion-induced thickness loss in ship hull structures through strain measurements obtained from in-situ sensors. This is introduced to regions where this phenomenon is common, and candidate sensor locations are selected based on the ship’s structural arrangement. Strain response data are obtained using a detailed FE model, which is subjected to a quasi-static loading condition, specifically still water loading. Operational variability is accounted for by considering the cargo filling rates as beta-distributed random variables. A Monte Carlo Simulation is utilized to estimate the statistical structure of the response, for both the intact and damaged conditions. Based on that, their distributions are considered mean-shifted and Gaussian. Additionally, the very low coefficient of variation of the monitored static strains allows for the measurements to be considered as deterministic signals. Therefore, the tools offered by classical detection theory of deterministic signals within noise are applicable and were ultimately used to construct a detector, whose performance was evaluated on an indicative 10% thickness loss scenario.

A Structural Health Monitoring (SHM) architecture involves the processing of sensor measurements and their translation to decisions about the structure’s condition. Given a specific SHM approach, the sensor topology, the feature selection, and the employed detector are the main elements that control the detection performance. This work provides an exploratory analysis of the statistical response patterns that govern a structure subjected to variable loads and methodically arrives at an optimal sensor topology, that maximizes the detection performance. For demonstration purposes, a thin square plate subjected to probabilistically described loads is considered. The damage of interest corresponds to a uniform thickness loss, the detection of which is evaluated at different damage levels (from 1% to a 90% unrealistic upper bound). The damage is to be identified indirectly, through strain sensing. The problem is numerically approached (Finite Elements and Monte Carlo Simulations). The generalized Gaussian likelihood ratio test is employed for setting up the detector. The effect of the feature vector arrangement to the detection performance is assessed through estimations of the probability of detection and false alarm, under the Neyman-Pearson framework. The optimal feature vector has been derived through case-based informal (selective process) or formal (Genetic Algorithms) optimization.

Modelling generic size feature of delamination, like area or length, has long been considered in the literature for damage prognosis in composites through specific models describing damage state evolution with load cycles or time. However, the delamination shape has never been considered, despite that it contains more damage information like the delamination area, center, and boundary for structural safety evaluation. In this context, this paper develops a novel particle filter-based framework for delamination shape prediction. To this end, the delamination image is discretized by a mesh, where control points are defined as intersections between the grid lines and the boundary of the delamination. A parametric data-driven function maps each point position as a function of the load cycles and is initially fitted on a sample test. Then, a particle filter is independently implemented for each node whereby to predict their future positions along the mesh lines, thus allowing delamination shape progression estimates. The new framework is demonstrated with reference to experimental tests of fatigue delamination growth in composite panels with ultrasonics C-scan monitoring.

Assessing the structural quality of metallic turbine blades is a challenging task due to their complex geometry and wide range of possible defect features. Process Compensated Resonance Testing (PCRT) is an effective quality assessment tool that uses a broadband sinusoidal swept input to excite the resonant modes of the component and employs a supervised learning algorithm to interpret the resonant modes and as such to determine the structural quality of the component. Our previous work has mostly centered on the exploitation of classifiers that are based on Mahalanobis distance. However, in practice, the measurement uncertainty may lead to bias in the trained classifier, potentially resulting in misclassification of the turbine blade.In this study, the concept of interval Mahalanobis space is introduced in the classifier in order to cope with measurement uncertainty. The resulting Integrated Interval Mahalanobis Classification System (IIMCS) classifier employs BPSO to screen those resonant frequencies that contribute favorably to the system and analyzes the sensitivity of resonant frequencies to measurement uncertainty under a Monte Carlo simulation scheme. The developed classifier was applied to an experimental case study of equiaxed nickel alloy first-stage turbine blades with a range of defect features, showing a good and robust classification performance.

For complex or large structures, the model updating process can be long and tedious and numerical methods can be computationally expensive. Hence, practitioners and researchers often resort to meta-modelling techniques when large problems are met. Even so, traditional methodologies, such as the Efficient Global Optimisation, can be slow and give sub-optimal results. This work proposes a new methodology for the model updating of numerical systems based on a novel Kriging approach for the scope of damage detection and quantification. The framework proposed is based on a global-local optimisation strategy recently developed by the authors, the refined Efficient Global Optimisation, herein used to tweak finite element models’ parameters to match the modal data extracted from a numerical system by using the residuals of the modified total modal assurance criterion. The main advantage to existing direct optimisation and meta-modelling frameworks is the more efficient use of computational effort for higher dimensional problems, which is verified with the use of a numerical system.

In this work, a novel statistical approach for damage detection and identification in the context of ultrasonic guided wave-based damage diagnosis is proposed using stochastic functional series time-varying autoregressive (FS-TAR) models. Wavelet functions are used as the functional basis family and the coefficients of projection of the time-varying model parameters are estimated via a maximum likelihood scheme. Damage detection and identification are tackled within a statistical decision making framework while appropriate thresholds are derived using pre-determined type I error probability levels. Both damage intersecting and non-intersecting, with respect to wave propagation, paths are considered in a multi-sensor aluminum plate in pitch-catch configuration. The method’s robustness, effectiveness, and limitations are discussed. The results indicate the effectiveness of the proposed method in detecting and identifying damage within a statistical setting.

Guided wave-based techniques have been used extensively in Structural Health Monitoring (SHM). Models using guided waves can provide information from both time and frequency domains to make themselves accurate and robust. Probabilistic SHM models, which have the ability to account for uncertainties, are developed when decision confidence intervals are of interest. However, most existing active-sensing guided-wave methods are based on the requirement that a relatively large data set can be collected, and thus are not feasible when data collection is restricted by time or environmental conditions. Meanwhile, simulation results, though lack of accuracy compared to real world data, are easier to obtain. In this context, models that incorporate data from different sources have the potential to embrace the accuracy of experimental data and the convenience of simulated data without the necessity of large and, potentially costly experimental, data sets. The goal of this work is to introduce and assess a probabilistic multi-fidelity Gaussian process regression framework for damage state estimation via the use of both experimental and simulated guided waves. The main differences from previous works include the combination of damage-sensitive features (damage indices; DIs) extracted from experimental and numerical sources, and a relatively small amount of real-world data. The proposed model is validated using data from two sources. The experimental data is collected from a piezoelectric sensor network attached to an aluminum plate under varying crack sizes while the simulated data comes from multi-physics finite element model (FEM) simulations with the same specifications.

The problem of random vibration response based robust and unsupervised damage detection for a population of composite aerostructures is addressed. The focus is on the achievement of robustness which is of paramount importance as manufacturing variability within the population and flight condition variability are practically inevitable. Two robust damage detection methods are postulated based on Multiple-Input Single-Output (MISO) Transmittance Function (TF) stochastic AutoRegressive with eXogenous pseudo-eXcitation (ARX) type representations for eliminating the effects of non-measurable excitation. Robustness to manufacturing variability is achieved via Multiple Model (MM) representations (the MM-TF-ARX method) or Principal Component Analysis (the PCA-TF-ARX method). The achievable detection performance is assessed via Monte Carlo ANSYS-based simulations with a population of 120 composite beams subject to manufacturing thickness variability, two distinct turbulence-like excitation profiles, and three early-stage crack damage scenarios. The results, in terms of Receiver Operating Characteristics curves, indicate excellent damage detection performance for the MM-TF-ARX method, yet inferior for its PCA-TF-ARX counterpart.

This study explores the feasibility for automated and robust detection of incipient faults in rotating machinery under different operating speeds using unsupervised vibration-based Statistical Time Series (STS) methods. The investigated faults cause no obvious effects on the time domain signals, while their effects on the signals power spectral density are almost completely masked by the effects due to the different operating speeds, leading thus to a highly challenging detection problem. Two unsupervised STS methods are employed, the Functional Model Based Method (FMBM) and a Multiple Model (MM) based one, while a single accelerometer is used on a rotating machinery that consists of two electric motors coupled via a claw clutch. The methods’ detection performance is assessed based on hundreds of experiments with the healthy machinery as well as with two incipient faults. One corresponds to slight wear at the base of a single tooth in the claw clutch spider and a second to tightening torque reduction at one of the four machinery mounting bolts, while it operates under 21 distinct speeds. The results indicate perfect detection performance via the FMBM overcoming that of the MM based method.

Structural Health Monitoring (SHM) via data-driven techniques can be based upon vibrations acquired by sensor networks. However, technical and economic reasons may prevent the deployment of pervasive sensor networks over civil structures, thus limiting their reliability in terms of damage detection. Moreover, the effects of environmental (and operational) variability may lead to false alarms. To address these challenges, a multi-stage machine learning (ML) method is here proposed by exploiting autoregressive (AR) spectra as damage-sensitive features. The proposed method is framed as follows: (i) computing the distances between different sets of the AR spectra via the log-spectral distance (LSD), providing also the training and test datasets; (ii) removing the potential environmental variability by an auto-associative artificial neural network (AANN), to set normalized training and test datasets; (iii) running a statistical analysis via the Mahalanobis-squared distance (MSD) for early damage detection. The effectiveness of the proposed approach is assessed in the case of limited vibration data for the laboratory truss structure known as the Wooden Bridge. Comparative studies show that the AR spectrum is a reliable feature, sensitive to damage even in the presence of a limited number of sensors in the network; additionally, the multi-stage ML methodology succeeds in early detecting damage under environmental variability.

The Inverse Finite Element method (iFEM), employing a network of strain sensors installed on a structure reconstructs the displacement field independently of the structural loading conditions and material properties. However, the solution is compromised when the sensor network, due to logistic or cost constraints, is sparse and measureless finite elements are present. To overcome this issue the iFEM minimizes a weighted functional, assigning smaller weights to the elements missing experimental measures. Strain field pre-extrapolation techniques have been proposed to improve the iFEM performance, although still assigning arbitrarily small weights to the extrapolated strains. The current paper proposes a Gaussian Process as the pre-extrapolation technique for the strain field, which natively incorporates measurement uncertainty, therefore providing a metric to assign the functional weights, as well as confidence intervals for the displacement field computed through the iFEM. The proposed technique is validated with a virtual experiment; advantages and limitations of the proposed approach are also discussed.

In monitoring system design, the expected uncertainty of key parameters representing the structural behavior must be smaller than a target uncertainty. As a result, measurements can be used to optimally manage civil infrastructure. An important key parameter in PC-bridges control is the trend over time of structural responses, such as strain, deflection, and rotation purged from temperature effects. In fact, temperature changes cause significant responses that we must filter out to isolate distortions due to rheological effects in concrete, strand relaxation, and long-term prestress losses. We can quantify the uncertainty of temperature-compensated key parameters a posteriori, after the acquisition of data. However, in the design phase, monitoring data are not available yet; therefore, we must estimate the expected uncertainty a pre-posteriori. In this paper, we propose a logical procedure based on Bayesian inference for evaluating the pre-posterior uncertainty of long-term measurements-trend purged from temperature effects. We verify to what extent this quantity depends on: (i) measurement and model uncertainties; (ii) expected monitoring duration; (iii) sampling frequency; (iv) seasonal temperature variation. We validate our approach on a real-life case study: the Colle Isarco viaduct, a box-girder prestressed concrete bridge.

CERN is the European Organization for Nuclear Research based in Geneva, Switzerland. There are over 300 underground structures, over 68 km in length, made up of tunnels, caverns, shafts and galleries, some of which were built over 60 years ago.During operational periods, exposure to ionising radiation make most of the underground infrastructure at CERN inaccessible for long periods, meaning that regular inspection and maintenance is a major challenge.CERN has recently invested significant effort in developing and implementing smart infrastructure and this is being used to overcome these challenges and enhance the way underground assets are managed. Technology at CERN allows the inspection and monitoring of priority infrastructure remotely, enabling regular inspections when access restrictions are in place. A focus on automation, using technology to unlock insights from inspection data, is being implemented so that a move towards a predictive, data-driven asset management system can be made, reducing the cost of repairs, and improving the safety of infrastructure.This paper will give a broad overview of the smart infrastructure concepts currently being developed at CERN. This includes fibre optical strain monitoring, robotics, computer learning fault detection, and digital survey tools using GIS and mobile applications.

The work presented in this paper is part of the LEVEL-UP project, which aim is to develop a platform for monitoring the production lifecycle of Large Industrial Equipment (LIE). One of these LIE involved in the project is cold stamping presses for automotive components manufacturing, for which, within the framework of this project, it has been developed a monitoring system based on two complementary Fiber Optic Sensor (FOS) technologies. A combination of fast and high accuracy punctual Fiber Bragg Grating sensors (FBG) with multipoint Brillouin distributed sensors were selected for the monitoring system. The multiparametric FOS monitoring system was installed and validated in the structure of three large cold stamping machines to analyze their Structural Health Monitoring (SHM), and for their Condition Monitoring (CM). During the validation, it was noticed that from the distributed sensors is possible to obtain a 3D map of the strain and temperature status of the press structure. And, from the FBGs, besides to obtain information from the strain and temperature of the points where these sensors are installed, also, they provide information about each stamping process: the applied force by the press, the frequency of manufacturing, and the number of components manufactured.

Monitoring and managing the health of critical structures has been identified as a key approach to define performance parameters in realizing the next generation of industrial manufacturing, production and operation (Industry 4.0 principles in Europe). However, stakeholders today are facing challenges in handling the data acquired from the sensors along with the complexities in post processing for improved productivity and safety standards. Although the currently available state-of-the-art sensors are capable of collection and communication of structural health related aspects, the complexity associated with the data processing scheme and effective wireless transmission of large data-volumes for real-time decision making hinders its value-add for industrial applications. In this context, we propose an early warning system, using smart multi-sensors, with the capability for on-board processing and data logging. Although the developed system could be applied for a wide variety of use-cases, across several industries (including aerospace, automotive and marine), we employ the same for monitoring the health of critical infrastructure elements in this study. Along with shorter decision making timelines, the developed solution would ensure increased cost savings due to a higher degree of automation, increased efficiency, robustness and safety of the critical structures and/or systems.

Transportation networks infrastructure ageing, deterioration and maintenance present a global problem and their loss translates to a loss of movement and transportation, trade and commerce, among others. The use of remote sensing technologies (non-invasive) is an efficient solution to enhancing the condition assessment of these from large-scale networks to single sections supporting improved decision making. Nowadays, the availability of a large amount of SAR data can be implemented in a Web-Based Spatial Decision Support System (WB-SDSS) platform to provide a powerful instrument for quick diagnosis to stakeholders. To this aim, a novel WB-SDSS solution has been developed in the Angular workspace. The IT architecture is based on PostgreSQL relational databases, periodically updated with new SAR data acquired and processed, and shared with users thanks to Geoserver services. The implemented Python code allows visualizing time series of Permanent Scatterers (PSs) obtained from free-of-charge SAR imagery in addition to data acquired by in-situ instruments (i.e. inclinometers, piezometers, etc.). Finally, to provide a useful instrument for quick diagnosis to stakeholders, employing the methodology of the I-Pro MONALISA algorithm presented by [1] the road network has been discretized in terms of areas with different degrees of attention.

The use of space-born data for monitoring single structures or infrastructures and detecting structural anomalies is recently becoming more widespread. The increasing application of satellite data in the field of Structural Health Monitoring (SHM) is due to the significant advantages that they should bring compared to the traditional monitoring techniques. In fact, satellite data can allow us to obtain information quickly not only on the structure of interest but even on the surrounding area. Satellite data can provide different kind of information based on their nature: interferometric satellite data can detect displacements with high accuracy; while other satellite data are able to provide geophysical information on the soil state (i.e., soil temperature, soil moisture, etc.). However, the interferometric data are generally applied separately from the other types of satellite data that are less used, so there is no combination at the moment in the SHM field. In this paper, an integrated use of different space-born data is proposed to understand if their cross-comparison can provide a more comprehensive overview of the structural behavior of a structure or infrastructure. As a benchmark, satellite data are applied to an ancient aqueduct, which is used to verify the proposed methods.

This work aims to investigate the potential of the application of Synthetic Aperture Radar Interferometry (InSAR) to bridge monitoring, in order to demonstrate its effectiveness in detecting damages early, preventing collapses and triggering in-situ inspections, especially if integrated with advanced numerical modeling. A post-processing methodology is presented, able to derive two-dimensional displacement configurations of multi-span bridges with properly defined error bounds, using both ascending and descending SAR acquisitions. The bounds are estimated by considering both random and systematic errors, accounting for the orientation of the bridge with respect to the Line-Of-Sights of the satellites, the hypothesized deformation plane and the accuracy of the measurements. The proposed procedure has been applied to the case study of the Albiano-Magra Bridge in Italy, collapsed on April 8th 2020. The results, referred to the monitoring period 2015–2020, highlight anomalies in the displacements of the first two spans and abutment of the East side, observed during the three years preceding the failure, thus providing information on the possible cause of collapse. A numerical model of the bridge, based on the Applied Element Method (AEM), was developed in order to simulate the progressive collapse of the structure subjected to the settlements derived by monitoring data and to verify the hypothesis on the kinematics and its causes.

Structural Health Monitoring (SHM) field gained increasing interest during the last years, also due to the huge amount of civil structures and infrastructures near or beyond their design life, and needs to be managed, particularly in a multi-hazard prone area like Italy. In this context, the advanced multi-temporal differential interferometry synthetic aperture radar (MT-DInSAR) techniques represent a great potential for the SHM applications and future development. The exploitation of such techniques in the monitoring and structural assessment of the built environment is still an open issue even if some first applications are available in literature showing promising potentialities. In this work, after a brief description of the general framework for structural monitoring and assessment, previously proposed by the authors, the case study of the Torri Stellari buildings in Rome (Italy) is presented and critically discussed. In particular, starting from COSMO-SkyMed (CSK) ascending and descending SAR datasets and applying the Small BAseline Subset (SBAS) DInSAR processing technique, the measurement points of the investigated area are obtained for both acquisition geometries. Different techniques are applied to both the displacement time series and mean deformation velocity for assessing the structural behavior or monitoring of constructions.

SAR Satellite monitoring is an interesting alternative to traditional Structural Health Monitoring (SHM) techniques since it does not require the installation of sensors and enables the collection of data on a large spatial scale. The data collected by satellites are processed to obtain time histories of displacements through which phenomena such as settlements at the base of a structure can be identified and followed over time.This article provides a general overview of SAR satellite monitoring of structures and infrastructures, with a specific focus on DInSAR techniques (“Differential Interferometry Synthetic Aperture Radar”). A data processing technique is presented to analyze DInSAR displacements to gain insight into the structural behavior. The case study investigated in this paper is the Palatino bridge in Rome, Italy. The SAR satellite data are acquired by COSMO-SkyMed of the Italian Space Agency and consist of displacements calculated during an acquisition period between 2010 and 2019. SAR satellite data allows to study the structural behavior in the vertical and longitudinal (i.e., along the bridge axis) directions. The effect of temperature on the bridge behavior is investigated.

It is crucial to monitor the wingtip deflection of aircraft in real-time to ensure its aerodynamic functionality and structural safety. Also, the prediction of the wing tip deflection can be highly useful to provide aircraft with improved maneuverability and agile response to sudden events, such as gust and flutter. In this study, we propose a wingtip deflection monitoring/prediction method based on machine learning techniques. Specifically, we employ an aerodynamic solver to simulate an aeroelastic flutter behavior of a fixed-wing, and demonstrate the feasibility of our approach to predict oscillatory wingtip motion due to the flutter. To track and predict the wingtip deflection, we build network architectures composed of convolutional neural networks (CNNs) to analyze optically measured data and recurrent neural networks (RNNs) to process time-series information. As a result, we find the proposed technique is capable of monitoring and predicting wingtip deflection in an accurate and efficient manner. We expect this vision and machine-learning-based technique can complement existing sensor technologies to enhance the safety and maneuverability of aircraft.

Aerospace, civil, energy, and mechanical engineering structures continue to be used despite reaching their design lifetime. Developing sensing and data analytics to assess the structural condition of the targeted systems is crucial. Traditional contact-based techniques may produce inconsistent results and are labor-intensive to be considered a valid alternative for monitoring large-scale structures such as bridges, large buildings, and wind turbines. Advancements in image-processing algorithms made techniques such as three-dimensional digital image correlation (3D-DIC), infrared thermography (IRT), motion magnification (MM), and structure from motion (SfM) appealing tools for structural health monitoring and non-destructive testing. Besides, as those techniques are implemented within unmanned aerial vehicles (UAVs), the measurement process is expedited while reducing interference with the targeted structure. This paper summarizes the research experience performed at the University of Massachusetts Lowell. The results of these activities show that the combination of autonomous flight with 3D-DIC, IRT, and SfM can provide precious insights into the structural conditions of the inspected systems while reducing downtime and costs. The study includes future research directions to make those approaches suitable for real-world applications.

Asset owners across the world are facing the challenge of ageing infrastructure. Critical assets such as bridges, tunnels, roads and sluices are approaching the end of their theoretical lifetime and the costs for renovation and replacement are rising rapidly. Asset owners are increasingly looking to make the transition from corrective or preventive maintenance to condition-based, predictive maintenance. By using a data-driven approach to asset management, the actual state of an asset can be understood better, maintenance can be optimized and the remaining lifetime estimated more accurately, whilst always guaranteeing the continued safety of the asset.The current methods for monitoring the behavior of civil infrastructure are not scalable to meet the needs of the industry in the coming decades. They either require labor-intensive manual inspections, or large numbers of electronic sensors involving complex, hazardous installation and maintenance procedures.Fugro has developed TotaLite®, a novel optical sensing system to overcome many of the issues associated with conventional deformation monitoring methods. TotaLite® combines innovative optics, smart image analysis algorithms and IoT data processing technology to deliver accurate contactless monitoring of displacements and deformations. Compared to existing deformation monitoring techniques, TotaLite® is easier to install, less invasive and uses less energy. It provides significant cost savings for existing projects, but will also make monitoring feasible for applications where currently no cost-effective methods exist at all.We present the measurement principles and data processing architecture underlying TotaLite®, as well as results of the first trials, comparing TotaLite® with existing monitoring techniques.

Damage detection often requires selecting a reference case to which examined instance is compared. This step is important as environmental and operational conditions should correspond in both cases. Commonly, selecting a reference case for damage detection in SHM is performed manually. As this approach is not effective, we introduce the Optimal Baseline Selection method based on analysis of frequency maps of examined structure. The operating object is recorded with the use of a high-speed camera. Vibrations are analyzed using the optical flow technique. The mean frequencies of motion in selected bands are calculated for each pixel. Frequency maps are obtained independently for vertical and horizontal directions of motion, and they are further processed separately. The analyzed frequency map is compared to the maps in a reference set to find the most similar one. This is basically achieved by computing differences between histograms of maps and further calculating statistical measures. Finally, the results obtained for individual directions of motion are combined. The method is demonstrated by an example of a compressor operating under different pressures and damages.

There are many methods for beam-like structures’ vision-based damage detection, usually based on deflection shape analysis. However, most engineering structures have geometries for which such a straightforward damage detection approach is not feasible. This paper introduces the Aligned Marker Space (AMS) method to close this gap and allow using straightforward deflection-difference-based damage detection methods for various civil engineering structures. The method involves transforming vision data into a new morphed space to straighten complex geometries along one dimension. Data transformed using this approach can be used as an input to various damage detection algorithms, including local (e.g., second-derivative-based or wavelet-based algorithms) and global approaches (e.g., trend detection and line-segments method). The method is verified based on data acquired in numerical simulations and in a set of practical experiments. A Blender-based approach is used to build plausible simulated data with a high resemblance to real-life examples. These examples are then validated using complex geometries in a series of physical experiments under laboratory conditions.

The evaluation of the structural integrity of components concerns different fields of engineering. The detection of defects is essential, especially in materials that could present discontinuities associated with the manufacturing processes, or after being mechanically loaded.The present work investigates the potential of full-field optical techniques to detect defects. These techniques could present an interesting alternative to specific techniques such as traditional ultrasound inspection or Laser Scanning Vibrometry (SLDV), which is very sensitive even at high frequencies. Specifically, 3D Digital Image Correlation (3D-DIC) is explored to detect inner defects. This full-field optical technique is based on measuring displacement and recent advances in camera resolution allows a high sensitivity.In this work, it is analysed different cantilever-beam specimens with different internal discontinuities. These specimens are made using additive manufacturing in order to control shape and size of discontinuities. Mode shapes are explored with 3D-DIC employing high-resolution cameras and a subsampling procedure to capture images. This procedure presents a high special resolution, unlike other approaches. Results are compared with those obtained by SLDV and simulations.This preliminary analysis shows the potential of full filed optical techniques in the determination of behaviours that indicates the presence of a defect.

Vision based techniques are successfully applied in for Structural Health Monitoring (SHM). Between them one can distinguish thermography, video-based methods and hyperspectral imaging method. Hyperspectral Imaging (HSI) is a method of obtaining an array of two-dimensional images over a wide range of wavelengths in the electromagnetic spectrum. HSI has found its applications in the fields of geographic remote sensing, food quality inspection, vegetation monitoring and medicine.One of the areas in which HSI usage is still developing is Structural Health Monitoring. There is a big potential of the HSI application because during hyperspectral imaging we can detect changes in physical and chemical properties of materials under the test. The aim of the study was to develop a technique for damage detection in glass fibre reinforced polymer composites, which is relatively difficult to achieve with other optical methods. The hyperspectral images for healthy and damaged composite samples are compared. For detection of damages, two approaches were investigated; a target detection algorithm using spectral data as the detection criterion, and an algorithm based on cointegration analysis. The strengths and weaknesses of both approaches are compared, and their applicability for SHM is assessed.

In the present study a comparison of frequently used computer vision (CV)-based methods for structural health monitoring of truss structures is shown. The attention is paid to template matching methods that can be classified into one of two groups: area-based and feature-based methods. Synthetic but realistic video is used in this study. Results of the comparison are reliable due to the fact that the exact displacements are known from the finite element model of the investigated structure. From the variety of tested CV methods, the Kanade–Lucas–Tomasi algorithm with FREAK-based repetitive correction outperforms the remaining tested methods in terms of the computation time with a negligibly greater estimation error.

Manual visual inspections typically conducted after an earthquake are high-risk, subjective, and time-consuming. Delays from inspections often exacerbate the social and economic impact of the disaster on affected communities. Rapid and autonomous inspection using images acquired from unmanned aerial vehicles offer the potential to reduce such delays. Indeed, a vast amount of research has been conducted toward developing automated vision-based methods to assess the health of infrastructure at the component and structure level. Most proposed methods typically rely on images of the damaged structure, but seldom consider how the images were acquired. To achieve autonomous inspections, methods must be evaluated in a comprehensive end-to-end manner, incorporating both data acquisition and data processing. In this paper, we leverage recent advances in computer generated imagery (CGI) to construct a 3D synthetic environment with a digital twin for simulation of post-earthquake inspections that allows for comprehensive evaluation and validation of autonomous inspection strategies. A critical issue is how to simulate and subsequently render the damage in the structure after an earthquake. To this end, a high-fidelity nonlinear finite element model is incorporated in the synthetic environment to provide a representation of earthquake-induced damage; this finite element model, combined with photo-realistic rendering of the damage, is termed herein a physics-based graphics models (PBGM). The 3D synthetic environment with PBGM as a digital twin provides a comprehensive end-to-end approach for development and validation of autonomous post-earthquake strategies using UAVs.

A robust structural health monitoring approach measuring the structural responses of bridges such as displacements, strains etc. helps to ensure their safety and serviceability. Static and dynamic loads from vehicles and pedestrians influence the instantaneous responses of bridges, while thermal loads from daily and seasonal temperature variations influence bridge long-term responses. Vision-based monitoring (VBM) is an emerging non-contact, non-destructive monitoring approach. It utilizes cameras to capture sequential images of the structure under load and suitable image processing algorithms for target tracking. VBM has shown promising accuracy in static and dynamic response measurements of bridges, however, the evidence of its accuracy in thermal response measurements is limited. This research illustrates the results of laboratory experiments implementing VBM for thermal response measurements. Thermal responses of a laboratory truss are monitored with VBM and contact sensors such as thermocouples and linear variable differential transformers (LVDT). Cyclic temperature loads are applied to the truss to simulate daily temperature variations. The truss is monitored with GoPro cameras and contact sensors. Measured response trends by VBM and LVDT are comparable, indicating the accuracy of VBM to measure thermal responses. Thermal responses measured by VBM are higher than those of LVDT, signifying requirement for measurement resolution enhancement. The measurement resolution of VBM is 0.099 mm/°C and LVDT1 is 0.041 mm/°C respectively. This discrepancy can be attributed to non-identical targets of VBM and LVDT, resolution of the camera, efficiency of the feature tracking algorithm and robustness of LVDT output. This case study illustrates the feasibility and challenges of VBM for thermal response measurement.

Distributed Optical Fiber Sensors (DOFS) are gaining interest for structural health monitoring applications. Recently, they have been successfully used to monitor the strain profiles at the interface between concrete and steel or composite rebars during pullout tests. In the framework of durability studies on Glass Fiber-Reinforced Polymer reinforcing bars (rebars), this method offers interesting perspectives for monitoring the evolution of the concrete/rebar bond behavior under accelerated ageing conditions. However, the durability of the bonded DOFS instrumentation itself should be investigated in a preliminary stage, to assess possible alteration of its performance over aging, which would raise questions about the validity of strain measurements in the long term.This article presents the first results of an experimental program aiming to address this specific issue. In this work, steel rebars were equipped with bonded optical fibers. Half-length of the instrumented rebars was subjected to hydrothermal ageing by immersion in an alkaline solution, while the remaining length was exposed to standard laboratory conditions. After exposure, the rebars were tested in tension and the DOFS strain profiles were simultaneously measured. These strain profiles were then compared to reference measurements performed before ageing, providing insights on the influence of the ageing conditions on the response of the DOFS.

In the Architecture, Engineering and Construction (AEC) industry, unmanned aerial vehicles (UAV) has been widely acknowledged as a promising tool to perform adaptive structural health monitoring automatically. However, there still remains some challenges for drones to collect image data of underground structures, primarily due to low light and no GPS conditions. In order to facilitate data acquisition, this article developed a mobile software development kit (MSDK) for drone using visual positioning and predefined controlling code, which enabled the drone to automatically fly along a designated sinusoidal route, whilst continuously taking videos and images of the tunnel surface. The developed MSDK was able to adjust the drone parameters (e.g., overlapping rate, inspection range, heading, flight direction between frames of the video) for different underground infrastructure conditions. Furthermore, a field test is conducted in an abandoned windless tunnel near Cork (Goggins Hill Tunnel) to test its feasibility. Results show that the 40-m difference between the designated routine and actual routine was 1.9%, and the collected data processed by Pix4Dmapper could reconstruct the complete tunnel scene and surface details. The navigation method proposed in this paper allows UAVs to perform automatic inspection without GPS, and the collected image data is used to build a tunnel panorama view.

Unmanned aerial vehicles (UAVs) equipped with imaging and laser sensors have shown their benefits in structural health inspection due to their aerial mobility, low cost, and efficiency. However, UAV applications in practice are limited by their level of automation, and man-piloted operation dominates to date. With vehicle automation on the horizon, autonomous structural inspection systems via robotic vehicles have become a possibility. Nonetheless, significant challenges exist for testing and validating in a physical environment. This paper proposes a virtual reality framework for developing autonomous UAV-based structural health inspection systems. The framework, built atop a gaming engine, implements algorithmic virtual sensing and control of a UAV that flies virtually in a complex built environment. In this paper, we test this framework with a virtual UAV with an open-loop control approach for structural health inspection, including waypoint-based control and simultaneous localization and mapping. We further discuss its full potential as an aerial robotics learning and validation platform for developing advanced data-enabled structural-space exploration, optimal control, and damage assessment.

The next generation of bridge structure status management system can be implemented by integrating small size unmanned aerial vehicle (UAV) to easily access some parts of the bridge and cut the cost of expensive equipment’s, 3D model reconstruction using Structure from Motion (SfM) photogrammetry, and Deep Learning methods for damage detection. These techniques were integrated in this study to construct a seamlessly enhanced bridge visual inspection system. A preliminary discussion was conducted to detect visual damages using DeepLabv3+ in a 2-dimensional bridge inspection images captured by the UAV video and incorporate it into the generated bridge 3D model. A benchmark case study was conducted for a two-span steel bridge with severe corrosion damage, and the result shows the bridge 3D model with corrosion damages. A mixed reality platform was also demonstrated to view and save the 3D model virtually which can be used for easily deterioration assessment and maintenance evaluation.

Thermography is a widespread technique employed to inspect different equipment through the analysis of measured temperatures. In this case, this method was used to calculate four figures of merit, which evaluate the level of wear in the refractory walls of a dedicated ferronickel electric smelting furnace. Thermograms were taken from 22-year-old refractory walls and new ones in stable state operation. Images were pre- and post-processed by means of self-developed algorithms, where filtering techniques were applied for noise reduction. Results showed higher values and dispersions for the old refractory wall and similarity in the distribution of results for both cases. The study indicates that thermography is a good non-destructive inspection method for following the evolution of wear behavior in refractories submitted to high rates of heat and erosion.

The current study discusses two aspects of an infrared thermography-based local SHM system for monitoring damage evolution during fatigue of composites materials, namely, pre-macrocrack damage (Stage I) followed by subsequent crack growth (Stage II).The crack density is a very well-known damage parameter representing the actual mechanical state of the material in terms of stiffness degradation. In effect, for laminates presenting off-axis laminae, crack density is useful for determining the “characteristic damage state” (CDS) that is related the load carrying capability of the laminate.In present research, a novel procedure is proposed for performing contactless measurements of crack density during static tensile tests by using temperature signal.The results have been critically compared with the strain waves signal acquired by acoustic emission sensors during the same tests. The proposed technique and procedure lead to estimate the crack density in those applications where it is difficult to detect transverse crack using a direct measurement from common experimental techniques.

The present work concerns the study of distributed defects on metal surfaces. A typical situation is represented by corrosion due to environmental effect, for example in pipings. In this case, the detection of both presence and corrosion amount inside piping is difficult and often requires the use of non-destructive techniques. Aim of this work is the feasibility analysis of active thermography for non-destructive detection and quantification of corrosion in steel sheets. In particular, the present work aims at developing a data processing technique for thermal results obtained on steel sound and corroder sheets by means of a laser thermal investigation. Both pulsed and lock in active thermography techniques are explored, in transmission configuration. The influence of excitation parameters on damage detection and quantification performance was also investigated. To this aim different samples of specimens were obtained with increasing corrosion damage, by means of keeping the specimens in a saline cell for increasing time intervals. For all the investigations, the reference signal consisted in the sound sheet thermal response. Data processing allowed to relate frequency and time parameters of the thermal response to the corrosion level. The entity of corrosion was also compared to the thickness specimen variation obtained by means of a metallographic optical microscope.

For composite materials, thermographic techniques have been successfully employed as a suitable monitoring procedure based on the thermo-elastic stress analysis, combined with dissipative effects in full-field data processing approaches for the rapid detection of critical stressed zones, where the delamination initiation or fiber cracks arise under cyclic loads, with the simultaneous aim to establish, at the earliest time, damage phenomena evolving and turning to significative magnitude.The present work briefly resumes different case of studies with fatigue characterization under tensile load of composite samples also in presence of notched holes, subjected to damage evolution in different ways, as function of load intensity and endurance. An interesting hybrid methodology for damage evaluation during cyclic loads is proposed through the combination of real-time thermographic recording with the consolidated and robust stiffness analysis (compliance analysis). Raw thermal and processed data are employed for a qualitative analysis of critical zones and proper damage indication before failure. In addition, proposed thermal parameters and experimental compliance data seem to indicate good correlation for damage detection in different cases.

Structural integrity of mechanical devices is of fundamental importance for reliability under the action of service loads. To properly design a mechanical device against fatigue failure, a long test campaign involving many specimens and time must be performed according to traditional fatigue tests protocol. However, fatigue is a very dissipative phenomenon in which a large amount of energy is dissipated in the surrounding environment. Moving from this assumption, the adoption of infrared thermography can dramatically decrease the amount of time to obtain reliable information regarding the fatigue life of materials and components. Risitano Thermographic Method (RTM) links the superficial temperature during a fatigue test with the dissipated energy for a given stress level. The whole fatigue life of a specimen is represented by an Energy Parameter, strictly dependent on the test frequency and stress ratio, and this allow to obtain, even with one specimen, the entire fatigue curve. The Static Thermographic Method (STM) allows to assess the first damage in a specimen subjected to static tensile test by monitoring the superficial temperature evolution. The obtained limit stress could be directly related with the onset of fatigue damage within the material if cyclically stressed. The aim of the present work is to investigate the relation between the energy release and the damage at different stress ratios within a stainless steel AISI 316L, both under static tensile and fatigue tests using RTM and STM. Moreover, microstructure analysis is carried out to identify possible failure sites.

An analytical framework was developed to correlate the intrinsic dissipation to the second harmonic of the temperature signal evaluated by means of the Fourier transform, in the case of tension-tension fatigue. The theoretical model assumes internal adiabatic conditions and that the elastic-plastic material obeys a Ramberg–Osgood law. It was found that the intrinsic dissipation is correlated to the second harmonic of the temperature signal by means of a parameter (the β parameter) that depends on the cyclic hardening exponent n′. Eventually, the theoretical model was applied to data obtained by step-wise fatigue tests carried out on plain specimens made of C45 normalized steel.

The present work proposes a frequency modulated Infrared NDT low-cost setup. In particular, heat modulation is human-controlled by acting on a simple switch used to trigger common halogen lamps. The acquired thermal sequence is processed by Matlab scripts based on Least Square Fitting (LSF) and Discrete Fourier Transform (LSF). LSF is in particular performed using an harmonic analytical model, where a nonlinear fitting procedure, based on a pattern-search algorithm, allows to optimise the harmonics frequencies. The proposed processing schemes are implemented on experimental data acquired from a Glass Fibre Reinforced plate with artificial back-hole and embedded patch defects. The two LSF and DFT approaches are compared investigating phase maps at different frequency bins.

The long term dynamic monitoring of the Santa Maria di Collemaggio basilica revealed an unexpected trend of the first natural frequencies decreasing from their initial values estimated in 2018. The decrement of the natural frequency could originate from several factors. The structural system derives from the arrangement of four building materials: masonry (predominant), reinforced concrete (RC), timber and steel. Masonry is unlikely to suffer rapid decaying of their mechanical properties. Reinforced concrete (RC) and the embedded steel should not present decaying phenomena in the first years after built. Likewise, the steel plates which connect the Cross-Laminated Timber (CLT) panels should not manifest a significant progression of corrosion in an indoor environment. Conversely, timber’s physical and mechanical properties, such as swelling and shrinkage, density, modulus of elasticity, strength have a time-dependent response, even in the long-term. Likely, the decaying of the natural frequencies depends on the CLT roof’s time-dependent behaviour due to the potential modification of its mechanical properties and boundary conditions (interaction with the steel plates, e.g.). This paper presents selected results from the basilica’s three-years continuous dynamic monitoring and discusses an elementary mechanical model representative of the dynamics transverse to the nave walls. The mechanical model described by a limited set of parameters drives the assessment of the CLT roof’s possible role in causing the detected decrement of the natural frequencies.

Since the early civil Structural Health Monitoring (SHM) applications, the need for effective tools to characterize the structural response under varying environmental conditions has been recognized as a critical factor for SHM reliability. As a consequence, the current structural condition as well as relevant historical data have to be taken into account to have an up-to-date representation of the actual physical system in operation. Digital twins can play a primary role in the evaluation of the current condition of the structure and in anomaly detection to support structural maintenance. Building Information Modeling (BIM) is currently showing a potential in this perspective thanks to the possibility of collecting miscellaneous data and information in a unified platform; however, its extension to facility management is an open issue, in particular when the digital models have to be complemented with data coming from SHM systems. So, an effective integration between the two is highly desirable. The present paper aims at demonstrating how relevant information coming from SHM systems can be integrated into a BIM model of a structure to make them available for interrogation and further analyses. The development of a digital twin of a historical structure combining SHM data, BIM, and finite element model updating is illustrated pointing out its promising applicative perspectives for structural maintenance.

In this work, the low velocity impact behavior of laminated polypropylene/plain woven basalt structures was investigated. In particular, given the extraordinary resistance of these structures to perforation, which has already emerged from previous investigations, the authors wanted to verify whether the possible presence of transverse junction points could further enhance this interesting mechanical behavior. However, in line with contradictory evidence already present in the literature regarding the use of seams, regardless of their arrangement, the impact results, collected with dart drop tests, highlighted the non-influence of stitches on the damage extent of the composite system examined.In particular, with the same impact energy, the load deflection curves are almost superimposable even if minimal differences have been found in terms of external plastic deformation (indentation depth) as the stitch configuration varies. Furthermore, the arrangement of through thickness constraints affects the extent of the delaminated area.

The vibration performance of pedestrian structures attracts the attention of several studies, especially with respect to unfavorable operational conditions or possible damage scenarios. Specific vibration comfort levels must be satisfied in addition to basic safety requirements, depending on the class of use, the structural typology and the materials. Careful consideration could be then needed not only at the design stage, but also during the service life of a given glass walkway or pedestrian system. Moreover, accidental events may require rapid diagnostic analyses with minimized efforts. In this paper, the attention is focused on the vibration analysis of in-service laminated glass slabs that are part of an indoor, case-study walkway. The attention is focused on modular units with identical geometry and restraints, but intact (LGU) or partially fractured (LGF) glass layers (as it was in August 2021). The in-field rapid analysis of classical performance indicators is thus addressed. Among others, dynamic identification techniques are applied for the analysis of the pre- and post-fracture vibration performances in order to quantify the effect of damage in glass. Parametric experimental outcomes are thus discussed in the paper, as obtained from Operational Modal Analysis under emergency experimental setup.

Monitoring large structures over their lifetime is becoming increasingly important as robust monitoring will ensure safety and financial benefits. Offshore wind turbines (OWT) are in need of continuous structural health monitoring (SHM) methodologies for scour detection. Reliable detection of scour requires robust selection of structural features to adequately represent the integrity of the OWT. Successful identification of operational modal parameters of OWT is challenging because of the effect of environmental and operational variability (EOV). Therefore, mitigating the effect of EOV is crucial to guarantee that the variability in the extracted features is caused by structural degradation. This work presents a methodology based on stochastic subspace identification enhanced by a clustering technique to identify stable and robust structural modes over time. Gaussian process regression is used to mitigate EOV due to its flexibility and non-parametric nature. Data collected from a control OWT and a scoured OWT in operation over 8-months is used for verification. The results have demonstrated that scour affects OWTs by reducing their natural frequencies. The results also indicate that scour induces OWTs to react differently to EOV. The outcomes of this work provide the basis for real time monitoring of OWTs, thus facilitating a reliable identification of scour.

The robustness of Non-Destructive Testing methods is typically quantified through the computation of Probability of Detection (POD) curves. Current design standards in aviation recommend the implementation of these curves also for Guided Waves based Structural Health Monitoring (GWSHM) systems. POD curves are greatly influenced by the uncertainties that may exist within the SHM acquisition due to operational and environmental conditions. These uncertainties play a particularly important role in the detection of Barely Visible Impact Damage (BVID) in complex composite structures. In this work, a digital clone platform that is created with the aid of a Bayesian calibration of a Finite Element (FE) model, is used to complement the experimental measurements. Based on the estimations of the platform it is possible to generate a numerical sample of impact events that allows the estimation of the Model Assisted Probability of Detection (MAPOD), considering the underlying uncertainties. The performance of the proposed framework is evaluated for a surface mounted network of sensors that is permanently attached onto a composite structure.

The functionality of bolted connections depends significantly on the conservation of the bolt’s preload, which entails the necessity to monitor these components. Traditional monitoring techniques using electro-mechanical impedance spectra are based on the observation of the structural resonances and antiresonances. The change of structural resonances is clearly mirrored for example in the resistance spectrum. This paper introduces a novel approach for such a monitoring process utilizing the susceptance. The susceptance is the imaginary part of the impedance’s reciprocal. In an experimental setup, where piezoelectric wafer active sensors have been attached to the bolt’s head, different preload stages of these components were set. It has been observed that the elastic mechanical deformation of the bolt’s head, which is caused by the preload, affects the piezoelectric wafer active sensor. As the properties of the sensor are well pronounced especially in the susceptance spectrum, a condition index based on the susceptance is proposed. Usually varying ambient temperature can occur in practical applications of bolted connections. It needs to be differentiated from the changes of the preload conditions in order to avoid false alarm of the monitoring system. Therefore, the proposed condition index has been additionally investigated in dependence on varying ambient temperature. The experiments show very promising results.

The present paper considers localization of stiffness-related damage in minimum-phase structural systems for which input and output data is collected both prior and posterior to damage formation. In an idealized setting, such damage induces a rank-deficient transfer matrix perturbation, whose null space carries information on the spatial damage distribution. The information can be materialized by applying the null vectors, referred to as dynamic damage-locating vectors (DDLVs), as loads to the given system, which will confine damage to the subdomain exhibiting rigid-body motion. In real systems, where the transfer matrix perturbation will not be rank-deficient, the DDLVs must be extracted from a quasi-null space, and this will impair the discrimination between undamaged and damaged subdomains. This paper explores the merit of adding flexibility in the design of the DDLVs by implementing virtual output feedback with the aim of ameliorating the discrimination problem. The virtual implementation allows one to design and realize closed-loop systems, and hereby closed-loop DDLVs, from the identified open-loop system. Different design criteria, each theoretically amenable to enhance the damage resolution provided by the closed-loop DDLVs, are proposed and subsequently tested in a numerical example.

In aviation, carbon-fibre reinforced polymers and titanium alloys are widely used materials. Adhesive bonds are a lightweight option to transfer loads between different materials, and failure prediction of such adhesive joints is of great importance. This research is concerned with static damage detection of lap shear specimens with Structural Health Monitoring methods. The specimen design is derived from single-lap shear test specimens according to ASTM D5868, with one joining partner made of additively manufactured Ti6Al4V and the other one made of a carbon-fibre reinforced polymer laminate. The adhesive layer consists of the matrix material of the carbon-fibre reinforced polymer laminate. In order to enhance the lap shear, the additively manufactured part is printed in a comb-shape and on a number of specimens, pins are printed to yield additional interlocking of the parts. For assessment purposes, several different Structural Health Monitoring methods are available. Of these, Acoustic Emission measurements are evaluated in detail for this study. When the crack extends during linear increasing the load, acoustic waves propagate through the joining partners. To enable their detection, a piezoelectric element is bonded onto the surface of the Ti6Al4V adherend.The Acoustic Emission results show hits which are well separated from a noise floor, facilitating the use of Acoustic Emission for this application. Damages can be detected earlier than by evaluating strain changes with a Digital Image Correlation system.

Measuring the electrical impedance of piezoelectric transducers is an important step in structural health monitoring (SHM) systems with damage detection based on the electromechanical impedance (EMI) technique. Studies reported in the literature usually analyze the performance of commercial and alternative measurement systems at low frequency. However, the excitation of the transducer with a high frequency, typically above 1 MHz, allows the detection of incipient damage, making it necessary to analyze measurement systems at high frequencies. As it is an analysis region still little explored for the EMI technique, it is necessary to analyze the repeatability of measurements, since the variation of the structural condition is given by the application of damage indices that quantify the differences between impedance signatures obtained before and after the damage incidence. Therefore, experimental tests were carried out using a piezoelectric diaphragm coupled to an aluminum structure, in order to verify the accuracy of measurements by varying the duration and amplitude of the excitation signals of a measurement system that uses sinusoidal signals. The results indicate that the variation of these parameters directly interferes with the precision of the measurements and the impedance signatures, enabling more reliable analyzes to be obtained about the structural condition.

In the context of recent reusable launch vehicles (RLV) developments, the SHM methods could be useful to help fast and economic revalidation of RLVs between launches, provided that the SHM transducers keep their functionality under extreme environmental conditions during flight.This paper focuses on the thermal cycling durability of PZT piezoelectric transducers bonded on a Carbon Fiber Reinforce Polymer (CFRP) composite plate. The temperature cycles are set to be representative of the thermal loadings that the structure of a launcher could experience under a thermal barrier: limited number of cycles (<10) of short duration (<10 min) and high temperatures for thermoset-matrix composites (150 $$^\circ $$ ∘ C). The heating is generated using a CO2 laser in order to obtain the steep thermal ramps (60 $$^\circ $$ ∘ C/min) and gradients through the plate thickness to best reproduce flight conditions. To our knowledge, very few research studies have been carried out so far on the resistance of SHM systems under such environment.The degradation of the system formed by the composite plate, the adhesive and the PZT transducer is studied, in comparison with the state of reference before cycling, with three different approaches. First, electromechanical impedance is measured during the entire thermal cycling. Then the functionality of a guided wave SHM system is evaluated based on the emission and reception of Lamb waves between PZTs attached to the surface of the plate. Finally, Laser vibrometry is used to provide a mapping of the emitted Lamb waves propagation and to identify possible debondings or changes of amplitudes of waveforms after the thermal cycles.

The present paper reviews a method based on output feedback eigenstructure assignment for locating damage in minimum-phase structural systems with measurable input and output. The methodological concept is to assign a subset of eigenpairs invariant to a postulated damage pattern in one structural subdomain at a time and then locate damage when the assigned eigenpairs, in an idealized setting, do not change due to damage. In real applications, where strict invariance is unattainable, damage is located when the discrepancy between the assigned eigenpairs from the reference and damaged configurations is minimized. The required output feedback is realized through offline signal processing of input-output data collected from open-loop testing, and the eigenstructure assignment is formulated with eigenvalues and left eigenvectors. The principles of the method are reviewed and subsequently demonstrated in the context of a numerical example.

Cantilever-suspended girder bridges are provided with intermediate hinges to reduce the effects of support settlement, including damage to hinge bearings caused by corrosion. To inspect, repair, and efficiently maintain these bridges, it is necessary to develop a monitoring technology. Accordingly, research have been conducted on a bridge response health-monitoring technology. In particular, the properties of bearings significantly influence the vibration characteristics of bridges. However, previous studies are yet to examine changes in bridge response due to the corrosion of hinge bearings. Using finite element analysis (FEA), this study aimed to investigate changes in the bridge response and modal parameters of cantilever-suspended girder bridges due to the corrosion damage levels of the hinge bearings. In the FEA, model-installed elastic spring was used shell elements. Furthermore, static and eigenvalue analyses were conducted. Static analysis demonstrated that the target bridge type exhibited more minor changes in its girder deflection due to bearing corrosion. In addition, the effects were quantitatively determined by altering the corrosion position of the hinge bearing in the bridge.

This study examines the performance of the refined time-reversal method (RTRM), which recommends best reconstruction frequency as probing frequency and uses an extended mode of the reconstructed signal for computing the damage/similarity index for damage detection under a varying thermal environment. The experiments were performed on a 3 mm aluminum plate with a block mass-type damage at three excitation frequencies: the best reconstruction frequency, the sweet spot frequency at which only $$S_0$$ S 0 mode is actuated, and an arbitrarily chosen frequency. Measurements were taken at two temperatures, 25 $$^\circ $$ ∘ C and 50 $$^\circ $$ ∘ C. The percent similarity between the reconstructed and original input signals showed a consistent decrease with the increase in damage height at the best reconstruction frequency. However, at the other excitation frequencies, the change in the similarity with the increase in damage height did not follow a consistent trend, which can lead to erroneous information about damage size. The percent similarity changes significantly with temperature rise at the sweet spot frequency and the arbitrarily chosen frequency. However, it showed minimal change with temperature rise at the best reconstruction frequency for undamaged and damaged states of different severity, making the RTRM suitable for baseline-free damage detection and sizing even under temperature variations.

Vibration-based damage detection approaches have received considerable attention in the field of structural health monitoring. Modal parameters are often adopted to define damage-sensitive features, since their strong physical meaning can help interpreting the structural condition. On the other hand, they are also sensitive to changes to environmental and operational conditions. This aspect is critical for an automatic damage detection, because changes in modal parameters due to varying external factors (e.g. temperature) can be greater than those caused by damage.In this context, this paper proposes an application to real data coming from long-term structural health monitoring of axially-loaded beam-like structures under realistic environment. These very common structural elements are usually subject to an axial load that changes under environmental and operating conditions. Since the axial load is not directly known in most real applications, assessing damage using modal-based damage features is a complicated task.In this paper, a data driven approach that does not require a knowledge of the axial load is proposed to filter out the environmental effects on modal-based damage features. The strategy has been successfully tested on data acquired in an uncontrolled environment, and resulted in being a promising solution for real structural health monitoring applications.

One of the major advancements in the field of civil engineering from the past few years is intelligent infrastructures. These structures will have the ability of sensing and will be accurately able to respond to the changes in the environment and external damages. To be able to keep up with the advancements we need to detect the damages to monitor the health of the structure. This research study aims to develop an efficient damage detection method using video of the damaged structure taken by mobile or camera mounted on a drone to replace the human visual inspection using Artificial Intelligence. For this system, an AI-based application is developed that will be able to recognize the damages accurately in the bounding box or targeted limiting area via a portable camera. An artificial neural network will be used to train the model and to classify the images obtained from video footage. As a case study and validation, a real damaged bridge video footage is considered.

A suitable combination of different methods and measurement techniques is often required for Structural Health Monitoring (SHM) of objects in situ. In the field of aerospace applications, the current research project “Combined acoustic and modal structure monitoring” is dealing with the development of a robust SHM system for damage identification in carbon-fiber-reinforced polymer (CFRP) structures under realistic and varying loads. The methods combined within the project are based on guided waves, acoustic emission, and different vibration monitoring techniques.The present paper deals with the application of the nullspace-based fault detection algorithm (NSFD) as a part of a holistic concept for the monitoring of an aircraft door surround structure. Due to the great sensitivity to changes in the statistical properties of the measured data, the algorithm reacts very sensitively to structural changes, but also to changes in the environmental conditions, such as changes in the external loads. In this paper, the impact of changing excitation conditions on the NSFD damage indicator is analyzed. Therefore, two different formulations for the NSFD algorithm are studied and compared concerning their sensitivity and robustness against changes of the external loads. Since the sensitivity and the results depend also on the algorithm set-up parameters, these are analyzed and taken into account. The evaluation of the results considers both, the robustness to the influence of the varying excitation and the sensitivity to damages in the analyzed structure.The theory and the algorithms are successfully tested with measured data sets from a CFRP-airplane structure.

Among the Structural Health Monitoring (SHM) systems, Guided Waves (GW) based ones have been increasingly investigated by both research community and industry since their reliability in damage detection, requirement of a lower power consumption and capability in the monitoring of extended areas with a reduced number of transducers. However, their use in real applications is still challenging. Among the critical aspects that can compromise the effectiveness of such SHM systems in the identification of structural anomalies, the operating conditions (temperature, loads, vibration, corrosion, etc.) require a particular attention. Damage detection algorithms and methods are thought mainly by referring to lower Technology Readiness Levels, typical of laboratory conditions where real scenarios cannot be easily reproduced. For this purpose, the Finite Element (FE) models find a key role for the understanding of the physics of governing phenomena driving the GW also in scenarios closer to the real one. In this paper, an experimentally validated FE model has been used to investigate on the effects of GW propagation mechanisms in a composite plate at 65 ℃. The effects of this in-service temperature have been highlighted by comparing GW dispersive behavior as well as polar plots with respect to the room temperature (20 ℃). According to the results, it has been observed a decrease of GW propagation velocity at 65 ℃.

Unknown inputs in nonlinear systems can be estimated using linear operations by treating the nonlinearity as pseudo-forces and adding these pseudo-forces to the unknown inputs. This paper examines the situation where the spatial distribution of the nonlinear terms is not known. The first issue is to determine the number of active nonlinearities, and this is done by noting that the number is equal to the effective rank of a certain matrix that can be formed using Fourier transforms of the outputs. If the number of unknown inputs plus the number of active nonlinearities is less than the number of outputs, and the spatial position of the nonlinearities and inputs is known, the problem is well-posed. In the case where the spatial distribution of the nonlinearities is not known but the matrix that describes it as pseudo-force is sparse, the problem can be usually solved by iterations without much difficulty, at least in modestly sized problems. When the matrix is not sparse, however, as is the case when the nonlinearities enter the system at coordinates that are not in the dynamic model, the solution is no longer unique.

Ultrasonic measurements have recently been applied in the context of structural health monitoring in civil engineering. The ultrasonic method is very sensitive to environmental conditions and material property changes, which might be caused by prestress losses and loading effects on the bridge. However, it remains a challenge to identify the damage mechanisms causing structural property changes from the above-mentioned factors influencing ultrasonics. The focus of this contribution is to investigate the correlation between ultrasonic velocities and temperature by the coda wave interferometry (CWI) analysis. Long-term monitoring of a prestressed concrete bridge model under field conditions was carried out to observe the effect of temperature. The homogeneous change in concrete caused by temperature was revealed by applying the CWI method. The threshold value of the best correlation coefficients for the linear regression of temperature and ultrasonic velocity changes was also studied. Experimental assessment will provide a ratio for temperature correction to evaluate the prestressing force decrease and its consequences on crack opening or extension.

The ‘Gänstorbrücke’ bridge between the cities of Ulm and Neu-Ulm is one of the best-monitored bridges all over Germany. In addition to an already active bride monitoring system, we have equipped the bridge with 30 ultrasonic transducers to explore the monitoring possibilities at an in-service large-scale reinforced concrete structure with continuous active ultrasonic measurements. The monitoring system is based on the detection of small changes in the entire signal, especially the multiply scattered parts of the recording, the so-called coda. Applying Coda Wave Interferometry (CWI), subtle changes in the signal can be detected and related to changing velocities in the area between source and receiver. A comparison of the results from coda wave interferometry with the strain measurements of the permanent monitoring system shows a correlation between strain measurements and CWI results. We discuss the challenges of changing environmental conditions, pose for interpretation of the results, and highlight the advantages of embedded versus externally attached ultrasonic transducers in permanent bridge monitoring, especially when coda wave interferometry is applied.

In earthquake-prone regions such as the United States’ Pacific Northwest, post-event damage assessment tools are needed to enable safe and speedy recovery missions. Currently, damage assessment is performed mainly by visual inspection and often impossible for structural members that are inaccessible, such as deep foundations or interior members hidden by cladding. This research explores the possibility of using embedded ultrasonic transducers to monitor reinforced concrete members for damage progression under reverse-cyclic loading. The proposed approach uses a combination of passive and active ultrasonic stress wave monitoring approaches. The former, commonly referred to as acoustic emission monitoring, captures the energy released from active fracture sources. The latter is based on traditional and coda ultrasonic stress wave monitoring, which can take advantage of the highly sensitive diffuse portion of recorded waveforms to capture minute changes in a material. In this paper we introduce select results from a full-scale column-footing subassembly that was tested in the laboratory using a reverse-cyclic lateral loading protocol representative of a subduction zone earthquake with varying axial loading. A combination of surface-mounted and embedded transducers was employed to monitor the specimen throughout the loading process. The findings demonstrate the complementary information that may be gained from the proposed combined monitoring approach to characterize damage progression in reinforced concrete members.

In this paper, we propose to monitor the evolution of damage in freeze-thaw cycles tests using embedded ultrasonic piezoelectric transducers. The transducers are used in emitter-receiver pairs. Concrete specimens were subjected to 70 freeze-thaw cycles of 12 h in temperature ranges from −20 ℃ to +17 ℃ and the monitoring system automatically and continuously recorded ultrasonic signals every 5 min. With such rich information, it is possible to monitor the difference between recorded signals during freezing and thawing phases, as well as transition times between these two phases. Several indicators were extracted from the recorded signals to evaluate the level of degradation of the concrete, including the traditional wave velocity and the first wave amplitude of the recorded signals as well as Coda wave parameters. The results show that the first wave amplitude is a more sensitive indicator of early freeze-thaw damage than the traditionally used wave velocity. Coda wave interferometry (CWI) also proves to be an interesting and complementary indicator to better understand the freeze-thaw damage mechanism in concrete.

In this paper, the authors investigate the energy harvesting in railway bridges. An analytical model is proposed for the estimation of the energy harvested from train-induced bridge vibration. The typology of the energy harvester studied in this work is a bimorph beam consisting of two piezoelectric patches on a substructure with a seismic mass. The energy harvested is tuned to find the optimum operating point for maximum power production. The optimization problem is constrained by the integrity of the substructure, and it is solved using a genetic algorithm. Finally, the feasibility of the proposed model is verified against an existing analytical solution and finite element model.

Damage detection algorithms employing the traditional acceleration measurements and the associated modal features may underperform due to the limited number of sensors used in the monitoring and the smoothing effect of spline functions used to increase the spatial resolution. This study presents a hybrid structural health monitoring method for vibration-based damage detection of bridge-type structures to overcome such problems. The proposed method is based on sensor fusion from conventional and computer vision-based acceleration measurements. Three commonly used damage measures are presented and employed: mode shape curvature method, modal strain energy method, and modal flexibility method. The accuracy of these algorithms with the conventional structural health monitoring approach is demonstrated in a numerical case study, where damage scenarios are simulated on a simply-supported bridge. The efficiency and accuracy of the proposed hybrid health monitoring methodology are demonstrated in a case study where the conventional acceleration measurements fail to detect and locate the damage. The outcomes of this study indicate a strong potential of the proposed method for damage detection.

The monitoring of the stability of bridges is a technical challenge that, in recent years, has been tackled by means of real aperture radar interferometers. Such technique extracts, for each radar range cell, the time series of the received signal phase, which is related to the displacement of the phase center of the radar range cell. By analyzing the displacement time series of the reflection points located on the monitored structure it is possible to extract the vibrational frequencies and amplitudes of such points. In this work we employ two different radar systems, working at 17 and 24 GHz respectively, to analyze the vibrations of a railway bridge in static conditions located in Galicia, North-Western part of Spain. The radar trials were realized by locating the radar systems at three different locations, one under the bridge with the radar Line-of-Sight parallel to the bridge longitudinal axis, one at the right and one at the left of the bridge. Such geometrical diversity allows to detect multiple vibrational modes, both on the vertical and tangential directions. Tangential and vertical modes are detected with amplitudes of tens microns. The results obtained using the two radar systems show a quite good agreement and are validated against accelerometer acquisitions, which is aimed at demonstrating the suitability of radar interferometers to measure bridge vibrations. The improvement of some performance of the radar survey using the higher frequencies radar is also discussed.

One of the distinctive characteristics of the Milan Cathedral is the presence of iron ties under the vaults of all 5 naves: those tension bars date back to the age of construction and still have an important role in supporting the lateral thrust exerted by vaults and arches.During the maintenance interventions preceding EXPO 2015, almost all the tie-rods of the cathedral underwent visual inspection, geometric characterization and hammer tests to evaluate their dynamic characteristics. After a brief summary of the experimental procedures, the paper focuses on presenting and exemplifying the structural identification approach adopted in the vibration-based assessment of the tensile force in the metallic ties. The presented procedure differs from the one proposed by Tullini and Laudiero (2008) since: (a) the Young’s modulus of the metallic material is assumed to be unknown and (b) the resonant frequencies of higher modes are used to solve the inverse problem.The application of the proposed procedure is mainly exemplified with reference to the tension bars exhibiting the higher stress level (larger than 100 MPa) as those elements are now permanently instrumented with vibrating wire extensometers. Selected results from the long-term monitoring of the tie-rods are briefly discussed as well.

An ultrasonic sonar-based ranging technique is introduced for measuring railroad tie 3D surface deflections. Tie deflection profiles can be used for flagging defective conditions of rail tracks such as center-binding, that can cause train derailments. The technique utilizes an array of air-coupled capacitive ultrasonic transducers oriented parallel to the tie with a minimum clearance of 3-in. from the rail to ensure contactless inspection. The proposed technique has the potential for detecting both the static and dynamic behavior of ties. The transducers are used in pulse-echo mode and the distance between the transducer to the tie surface is computed by tracking the time-of-flight of the reflected wave from the tie surface. An adaptive, reference-based, cross-correlation operation is used for measuring relative tie deflections. Multiple measurements along the width of the tie allow the measurement of twisting deformations as well. An acoustic signal strength-based and an image-based classifier are used in tandem to differentiate signals coming from ties and ballast. Data acquisition and processing are performed using a LabVIEW real-time module. Field tests on a replica test track, at walking speeds, were performed at the Rail Defect Testing Facility. The initial results indicate the potential of this system to measure full-field tie deflections in 3D at walking speeds (~3 mph).

Additive manufacturing (AM) processes are rapidly maturing and being adopted in numerous industrial sectors. One of the big challenges with many AM processes is the need for part quality control, either in post-manufactured assessment or in-situ during the build. This paper presents a low-cost structured light system (using camera and projector) that exploits digital fringe projection to achieve surface profiling of AM parts. Additionally, a probability density function of the surface profile is derived, helping the measurement process to provide the probabilistic support required for AM part quality control decisions. Results from a prototype system on AM parts are demonstrated.

Distributed Optical Fiber Sensors (DOFS) show several inherent benefits with respect to conventional strain-sensing technologies and represent a key technology for Structural Health Monitoring (SHM). Despite the solid motivation behind DOFS-based SHM systems, their implementation for real-time structural assessment is still unsatisfactory outside academia. One of the main reasons is the lack of rigorous methodologies for uncertainty quantification, which hinders the performance assessment of the monitoring system. The concept of Probability of Detection (POD) should function as the guiding light in this process, but precautions must be taken to apply this concept to SHM, as it has been originally developed for Non-Destructive Evaluation techniques. Although DOFS have been the object of numerous studies, a well-established methodology for their performance evaluation in terms of PODs is still missing. In the present work, the concept of Probability of Delamination Detection (POD2) is proposed for a DOFS network; Carbon Fiber-Reinforced Polymers (CFRP) Double-Cantilever Beam (DCB) specimens equipped with DOFS have been tested under static loading, and the strain patterns along with the relative observed delamination size have been collected to generate an adequate database for the POD analysis, suggesting a reference methodology to quantify the performance of DOFS for delamination detection.

Synchronization discrepancies when measuring structural response may compromise the accuracy of information on the condition of structures, extracted from structural health monitoring (SHM) systems. Loss of synchronization mostly concerns modern wireless SHM systems; however, synchronization discrepancies may also occur in cable-based systems in case multiple data acquisition units without global clock management are used. This paper presents the implementation and validation of a concept towards synchronizing SHM data records containing structural response data, which is imperative for obtaining information valuable for accomplishing monitoring objectives. Specifically, the performance of a cross spectrum synchronization approach reported in a recent study, which builds upon cross spectral density phase angles of structural response data transformed into the frequency domain, is investigated with structural response data from real structures. The cross spectrum synchronization approach is implemented into a wireless SHM system and into a cable-based SHM system, and it is validated through tests on a laboratory shear-frame structure and through field tests on a pedestrian bridge, respectively. The validation test results prove that the cross spectrum synchronization approach is capable of complementing traditional clock synchronization protocols, thus enhancing the accuracy of SHM outcomes.

Reliability assessment of Structural Health Monitoring (SHM) systems poses new challenges pushing the research community to address many questions which are still open. For guided wave-based SHM it is not possible to evaluate the system performance without taking into account the target structure and applied system parameters. This range of variables would result in countless measurements. Factors like environmental conditions, structural dependencies and wave characteristics demand novel solutions for performance analysis of SHM systems compared to those relying on classical non-destructive evaluation. Such novel approaches typically require model-assisted investigations which may not only help to explain and understand performance assessment results but also enable complete studies without costly experiments. Within this contribution, a multi input multi output approach using a sparse transducer array permanently installed on a composite structure to excite and sense guided waves is considered. Firstly, the method and the analysis of path-based performance assessment are presented considering an open-access dataset from the Open Guided Wave platform. Then, a performance analysis of a guided wave-based SHM system using Probability of Detection is presented. To explain some unexpected results, the model-assisted investigations are used to understand the physical phenomena of wave propagation in the test specimen including the interaction with damage. Finally, issues and future steps in SHM systems’ performance assessment and their development are discussed.

Sensor network design and reliability assessment are critical aspects to achieve an effective guided wave based SHM system especially in composite materials. The detection capability of a SHM system exceedingly depends on its sensor network layout. Besides, Probability of Detection (POD) method is most common assessment method among few in use which requires vast amount of experimental data for each sensor network layout. Due to impracticality of repetitive experimenting, Model-Assisted Probability of Detection (MAPOD) applications attract interest recently with the help of efficient simulation tools. However, MAPOD studies requires thousands of simulations for each sensor network placement setup as well. In this contribution, to form a sensor placement while decreasing the simulation effort, a method called Sensitivity Analysis is proposed. The methods are based upon the idea of measuring maximum energy density coverage over the specimen to evaluate higher possibility of damage interaction while solving a high frequency approximation or energy accumulation with minimal cost, respectively. Demonstration case will be evaluated for an anisotropic composite component. The paper presents determination of sensor layout and validation with the comparison of well studied MAPOD simulations by using the Elastodynamic Finite Integration Technique (EFIT).

Performance assessment for GuidedWave (GW)-based Structural Health Monitoring (SHM) systems is of major importance for industrial deployment. With conventional feature extraction methods like damage indices, path-based probability of detection (POD) analysis can be realized. To achieve reliability quantification enough data needs to be available, which is rarely the case. Alternatives like methods for performance assessment on system level are still in development and in a discussion phase. In this contribution, POD results using an Artificial Intelligence (AI)-based data analysis are compared with those delivered by conventional data analysis. Using an open-access dataset from Open Guided Wave platform, the possibility of performance assessment for GW-based SHM systems using AI-based data analysis is shown in detail. An artificial neural network (ANN) classifier is trained to detect artificial damage in a stiffened CFRP plate. As input for the ANN, classical damage indicators are used. The ANN is tested to detect damage at another position, whose inspection data were not previously used in training. The findings show very high detection capabilities without sorting any specific path but only having a global view of current damage metrics. The systematic evaluation of the ANN predictions with respect to specific damage sizes allows to compute a probability of correct identification versus flaw dimension, somehow equivalent to and compared with the results achieved through classic path-based POD analysis. Also, sensitive paths are detected by ANN predictions allowing for evaluation of maximal distances between path and damage position. Finally, it is shown that the prediction performance of the ANN can be improved significantly by combining different damage indicators as inputs.

Population-based Structural Health Monitoring (PBSHM), has emerged in the recent past as a means of creating diagnostic systems across populations (or fleets) of structures. This idea represents a powerful approach to solving the problem of data-scarcity in conventional SHM, where data are not available for all health states of interest. A critical element of PBSHM is a means of assessing the similarity of structures or sub-structures, so that knowledge is only transferred between structures if they are appropriately similar. One similarity measure has been developed on the basis of irreducible element models and graph theory. This idea was illustrated in the first part of this study, using highly-simplified “schematics” of structures – bridges and aeroplanes. The objective of the current paper is to show how the methodology applies to complex real structures, including a Hawk T.Mk1 jet aeroplane and suspension bridges.

To move towards a condition-based maintenance practice for aircraft structures, design of reliable health management methodologies is required. Development of diagnostic methodologies is commonly realised on simplified sample structures with assumptions that methodologies can be adapted for application to realistic aircraft structures under in-service conditions. Yet such actual applications are not conducted. In this work, we study the development of diagnostic methodologies to training structures and their application to dissimilar testing structures. A heterogeneous population is considered, consisting of single-stiffener composite panels for methodology development and training and a multi-stiffener composite panel for application and testing. Characteristics as its composite material, lay-up, and temperature condition are constant while topologies and applied loads differ between the dissimilar structures. Damage in the structural panels is monitored on multiple diagnostic levels using a variety of structural health monitoring (SHM) techniques, including acoustic emission and distributed strain sensing. Specifically, we develop diagnostic methods for localising and monitoring disbond growth after impact using strain data collected during fatigue testing of multiple single-stiffener panels and apply these for disbond monitoring in an upscaled version of a multi-stiffener panel. In this manner, this study aids in the maturement and application of SHM methodologies to realistic aircraft structures.

Modelling the times-to-failures in industrial assets is critical for their maintenance planning. Modern fleets often comprise of clusters of similarly deteriorating assets, due to customisation or diversity in their operational settings. As such, the prevalent techniques of using a single fleet-wide model or independent cluster-specific models for modelling the times-to-failures are associated with high bias or high variance respectively. The problem of high variance is especially prominent for the asset clusters with where relatively fewer failures are observed. This paper proposes that statistical hierarchical modelling systematically mitigates the problem of high variance for the clusters with sparse data using shared higher level models for the cluster-specific model parameters. A hierarchical model of the Weibull density functions is shown in this chapter as an example because of the popularity of the Weibull distributions in reliability applications. Failure trajectories from a fleet of simulated turbofans are used herewith for the experiments that compare the advantage of the proposed hierarchical model over independent or fleet-wide models for modelling the observed times-to-failures.

Population-based structural health monitoring (PBSHM), expands the implementation of structural health monitoring concepts from a single structure to a group of structures. Within the populations of interest, it is useful to assess the similarity of structures, in order to form communities and networks; i.e., to establish clusters. By doing so, it is possible to infer whether transfer learning is likely to be applicable across a population. To address this, structures are represented in graphical form via irreducible element (IE) models or attributed graphs (AG), that encompass the topology, material properties and geometry of the structures. Kernel-based methods – also known as graph kernels – can then be applied to these IEs and AGs to assess the similarities within the population.In this paper, a number of comparisons are made between graphical representations of structures at different levels of resolution; these vary from assessing similarity at a purely topological level, to an increase in complexity, where discrete and continuous node labels are evaluated via graph kernels.

The structural health monitoring and assessment tasks of civil infrastructures e.g. bridges, railway tracks are inevitable in order to keep the transportation network active for smooth operation. And due to the complex inherent traits of the aforementioned infrastructures, it requires appropriate strategies to keep them functional. The conventional monitoring approaches (e.g. manual) are becoming unpopular as a result of the advancement of the sensors-based monitoring. Based on the measured data, long-term monitoring is possible regardless the existence of a prior model, though it may not be so straightforward task if a prior model is missing. Therefore, to avail the advantages of the state-of-the-art tools, this study focuses into the possibility of data-based modelling by adopting sub-space method. To do this end, two different experimentally measured sets of data of (i) a laboratory scaled bridge, and (ii) a rail-track’s data have been used to perform analyses. Initially, the performance of the developed models are evaluated and later the models have been validated. The pilot outcome shows the efficacy of the developed model in terms of prediction capability to the measured data. Further, the model has been utilized to forecast future behaviour that will assist the assessment of the unseen future behaviour of the infrastructures.

Vibration-based structural health monitoring represents an efficient way to evaluate structural integrity and the presence of damage at an early stage. These methods usually assume that damage manifests itself as a deviation in the modal properties of the structure with respect to its normal conditions. Traditional procedures for modal parameters estimation require the use of a dense sensor arrangement and complex logic techniques, thus making them not particularly suitable for the case of large engineering structures, where the need for cost-effective monitoring solutions is of utmost importance because of the large number of substructures to be monitored. This paper proposes the use of simple statistical and spectral features as a mean to characterize accelerations signals. Starting from this set of features, the principal component analysis (PCA) is first used to reduce data dimensionality still preserving the relevant information about the structural conditions, then a k-Nearest Neighbors (k-NN) procedure is adopted as a supervised machine learning method to classify different types of damage. The procedure is validated using the experimental data from the permanent monitoring system of the G. Meazza stadium grandstands of, where one accelerometer per stand is installed to get vibration data during the main events. Four grandstands located on the same ring and having the same nominal geometry are considered. The leading idea is to reproduce different scenarios where, due to the impossibility of imposing realistic damages, one grandstand is assumed to be the safe structure, while the others represent a proxy for small structural changes to be identified.

Concrete has evolved to be the most used construction material all over the world from minor to major structures. The accelerated deterioration of the structure can be reduced if the cracks are detected at their early stage, and it will be more effective if the crack type is also known. Development of crack and its propagation if unnoticed overtime can lead to minor to major damage. The proposed study targets to revolutionize the visual inspection process using Deep Learning (DL) algorithms. The most important part of training a DL algorithm is the data, hence the major challenge for the study is collection of various images consisting different types of cracks from the visual inspections carried manually during real-time projects. Overall, 3500 images were collected and out of which considering the computational complexity the proposed model was developed based on a 500-image sub dataset. The pre-trained DL model has demonstrated the desirable accuracies while no pre-trained model has dropped the accuracy. In practice, the paper also focuses on the current achievements and limitations of the existing methods and as well as opens lot of avenues for the upcoming researchers in the cross disciplinary fields of Structural and Computer Science.

Since 1988 onwards the VEGA-A platform, a steel jacket offshore structure operating in the Mediterranean sea near Siracusa in Sicily, is equipped by a monitoring system collecting environmental data (intensity and direction of wind, sea wave and sea direction) and structural response data (main frequencies and mode shapes). For these structures, in which the major part of the components is in the deep-water, analysis of the collected structural data and their cross-validation with the meteo-marine ones plays an important role in the maintenance schedules. In fact, a modification of the main frequencies and corresponding mode shapes over time may be related to a structural damage so, the capability to correlate structural and meteo-marine data allows a correct planning of the yearly maintenance operations. After a description of the monitoring system installed on the VEGA-A platform, this paper reports a statistical interpretation of the structural response over about 30 years of activity. The dynamic response of the platform under different sea conditions over these years is analyzed, focusing on the correlation between the dynamic parameters and the corresponding atmospheric and sea conditions.

Structural Health Monitoring (SHM) of ageing infrastructures represents one of the most important and challenging issues in modern society. In this framework, the increasing number of bridge collapses has progressively fueled the interest towards the development of reliable monitoring strategies, able to efficiently ensure real-time bridge assessment and early-stage damage detection. To this aim, recent advancements in sensor technologies and data science have strongly encouraged the use of Machine Learning (ML) algorithms. Within the unsupervised learning context, this paper proposes an innovative convolutional autoencoder-based damage detection technique applied to the Z24 benchmark bridge. Firstly, raw acceleration sequences of user-defined length are converted into images using the Gramian Angular Field Approach (GAF). Then, during the training, the model learns how to correctly reconstruct healthy data acquired by SHM systems. The trained network is afterwards tested with unknown data and the mean absolute error (MAE), considered as a damage-sensitive feature, is adopted to quantify the errors between the original input and reconstructed output. Finally, damage detection is performed when the percentage of damaged sequences within a pre-defined macro-sequence exceeds a properly fixed threshold. Results prove that the implemented ML algorithm, working at the level of single sensor, is effective to enable continuous assessment of large-scale monitored bridges as new data is collected, involving limited computational efforts.

The idea of exploiting sparseness in under-determined damage characterization problems is not new, and regularizations techniques that tend to promote sparseness, such as L1-norm minimization, have been investigated in the last ten years or so. Although various claims of merit have been made, two interconnected issues put these claims into question, and this paper brings some attention to the matter. The first is that the relationship between the structural parameters and the modal features previously considered has been linear and to ensure that the premise was closely realized, only very small damage severities have been considered. The second issue, intimately related to the first, is the fact that the noise, which has been typically taken as small relative to the “change in the features”, is then unrealistically small. In problems where the damage is sufficiently large, the nonlinear dependence of the features on the parameters cannot be generally discarded. It is found that the attainable performance is much less “impressive” that what has been often claimed. The paper also examines the potential merit of using Lp-norm (0 < p < 1) minimization, instead of L1-norm minimization which, to the knowledge of the writers, has not been previously examined in damage characterization research. In this case we also find that, contrary to claims made in other areas, this norm does not lead to any general improvement over the performance attained by minimizing the L1-norm.

The growing need for assessing the integrity of aging monumental structures by means of cost-effective and not-destructive techniques has driven significant scientific interest in vibration-based structural health monitoring (SHM) procedures, allowing to track changes in selected damage-sensitive structural parameters. However, the evaluation of a healthy or damaged state from the acquired monitoring data is often a data-driven process that can be subjected to a large amount of uncertainty, resulting in false positives and false negatives. Hence, the main idea behind this work focuses on handling the main issues related to the uncertainty management by exploiting the aggregation of different sources of information. In this context, this study is aimed at detecting and locating structural damages in monumental structures with the aid of a data fusion approach including vibration-based system identification, Bayesian-based finite element (FE) model updating and visual inspections. As a preliminary step, potential damage-sensitive sections are defined on the basis of nonlinear static analyses (NLSA) performed on a calibrated FE model and/or engineering judgment (EJ). Then, a surrogate model is established enabling to transfer knowledge from the monitoring data to the updated numerical models and to solve the inverse problem aimed at deriving the posterior statistics of the uncertain parameters. The effectiveness of the proposed approach is demonstrated by using 1-year of recorded data acquired in a monumental structure named Consoli Palace, located in Umbria, central Italy, a region characterized by high seismic hazard. The palace has been continuously monitored by the Authors since 2017 using dynamic, static and environmental sensors.

This paper presents a methodology to move toward reliable real-time structural health monitoring (SHM). The proposed procedure relies upon surrogate modeling based on a multi-fidelity (MF) deep neural network (DNN), conceived to map damage and operational parameters onto sensor recordings. Within a stochastic framework, the MF-DNN is adopted by a Markov chain Monte Carlo sampling procedure to update the probability distribution of the structural state, conditioned on noisy observations. The MF-DNN enables to locate and possibly quantify the presence of damage, and its multi-fidelity configuration effectively blends datasets featuring different fidelities without any prior assumption. The training datasets are generated with physics-based models of the monitored structure: high fidelity (HF) and low fidelity (LF) models are considered to simulate the structural response under varying operational conditions, respectively in the presence or absence of a structural damage. The MF-DNN is a composition of a fully-connected LF-DNN, which mimics sensor recordings in the undamaged condition, and of a long short-term memory HF-DNN, which is exploited to enrich the LF approximation for the considered damaged scenarios. By framing the model updating strategy as an incremental or residual modeling problem, the MF-DNN is reported to provide numerous advantages over single-fidelity based models for SHM purposes.

SHM methods based on finite element (FE) model updating are frequently used and have become popular because of its ability to estimate unknown system parameters by matching the predicted behavior to the observed structural behavior, which can often be measured under operational conditions However, although FE model updating methods have been usually applied for damage identification purposes, their application to large-scale redundant and complex structures is a difficult task due to the large number of degrees of freedom and the high number of unknown parameters to be updated, which involves a high computational cost and can drive to an ill-conditioned identification procedure and non-uniqueness of the solution. A novel localised damage function method called roaming damage method (RDM) capable of identifying individual members is proposed in this work. This method has the ability to identify a wide range of damage types, from large areas of low damage to individual beams which have been severely damaged.

The paper studies the topological changes from before and after cointegration, for the natural frequencies of the Z24 Bridge. The second natural frequency is known to be nonlinear in temperature, and this will serve as the main focal point of this work. Cointegration is a method of normalising time series data with respect to one another - often strongly-correlated time series. Cointegration is used in this paper to remove effects from Environmental and Operational Variations, by cointegrating the first four natural frequencies for the Z24 Bridge data. The temperature effects on the natural frequency data are clearly visible within the data, and it is desirable, for the purposes of structural health monitoring, that these effects are removed. The univariate time series are embedded in higher-dimensional space, such that interesting topologies are formed. Topological data analysis is used to analyse the raw time series, and the cointegrated equivalents. A standard topological data analysis pipeline is enacted, where simplicial complexes are constructed from the embedded point clouds. Topological properties are then calculated from the simplicial complexes; such as the persistent homology. The persistent homology is then analysed, to determine the topological structure of all the time series.

This contribution presents, discusses and illustrates the constrained mode decomposition (CMD) method. The CMD is a recently proposed method that extracts single mode components from measured multimodal free structural responses. These components can be then processed, in time domain or in frequency domain, for identification of modal parameters, and ultimately, for structural health monitoring. The aim of the CMD is thus similar to the aims of other well-known mode decomposition approaches, such as the empirical mode decomposition (EMD) or the variational mode decomposition (VMD). However, in contrast to the EMD, the CMD-processed responses retain the characteristics of the free response (satisfy the equation of motion of the same structure) and they have thus a clear, well-defined physical meaning. In comparison to the VMD, the formulation of the CMD is much simpler: the CMD combines linearly recorded structural responses in a way that simultaneously (1) amplifies the selected modal component and (2) constrains/suppresses other components. The amplification/suppression process is quantified in terms of the FRF peaks or, in case of closely spaced modes, in terms of FRF derivatives.

Average daily traffic (ADT) is one of the parameters that assesses the average traffic that runs on a bridge each day. ADT impacts the performance of every bridge in the long run. Deterioration rate of various bridge decks constructed with different materials in several climatic regions is different. Understanding the impact of ADT on various bridge decks built in different climatic regions allows the bridge owners to monitor the daily traffic closely and to efficiently allocate their funds for the rehabilitation of bridges. This paper attempts to study the impact of ADTs by calculating the deterioration rates of a variety of bridge decks that are made up of diverse materials and built in several climatic regions. Non-linear regression equations are developed using the time in condition rating mechanism to see how fast the bridge decks decline from an excellent condition to a satisfactory condition. Results show that the decks with lower ADTs that built with concrete continuous material and reside in Midwestern region deteriorate faster than the lower ADT bridges that were made up of concrete material and reside in Southeast region.