Search

Smart structural health monitoring (SHM) for large-scale infrastructure is an intriguing subject for engineering communities thanks to its significant advantages such as timely damage detection, optimal maintenance strategy, and reduced resource requirement. Yet, it is a challenging topic as it requires handling a large amount of collected sensors data continuously, which is inevitably contaminated by random noises. Therefore, this study developed a practical end-to-end framework that makes use of physical features embedded in raw data and an elaborated hybrid deep learning model, namely 1-DCNN-LSTM, featuring two algorithms—convolutional neural network (CNN) and long-short term memory (LSTM). In order to extract relevant features from sensory data, the method combines various signal processing techniques such as the autoregressive model, discrete wavelet transform, and empirical mode decomposition. The hybrid deep learning 1-DCNN-LSTM is designed based on the CNN’s capacity of capturing local information and the LSTM network’s prominent ability to learn long-term dependencies. Through three case studies involving both experimental and synthetic data sets, it is demonstrated that the proposed approach achieves highly accurate damage detection, as accurate as the powerful 2-D CNN, but with a lower time and memory complexity, making it suitable for real-time SHM. Note to Practitioners —This article aims to develop a practical data-driven method for automatically monitoring the operational state of structures. In order to achieve consistently and highly accurate results in performing different tasks for diverse structures, we combine underlying features in both time and frequency domains extracted from measured signal vibration data. Three popular data featuring methods are combined to achieve the diversity gain which would not be possible with each individual method. As the vibration is usually measured by long time-series signals, the most efficient deep learning architecture for time-series signal, namely long-short term memory (LSTM), is considered for this work. Besides, each structure has its own dynamic properties, i.e., eigenfrequencies, around which the most relevant information is in the frequency domain, thus convolutional neural network specifically designed for capturing local information is used in combination with LSTM, forming a hybrid deep learning architecture. The applicability and effectiveness of the proposed approach are supported by three case studies with different types of structures, showing highly accurate damage detection with reduced resource requirements. These advantages can be valuable for developing a model for live monitoring of structural health in the future life-line infrastructures.

Over recent decades, the artificial neural networks (ANNs) have been applied as an effective approach for detecting damage in construction materials. However, to achieve a superior result of defect identification, they have to overcome some shortcomings, for instance slow convergence or stagnancy in local minima. Therefore, optimization algorithms with a global search ability are used to enhance ANNs, i.e. to increase the rate of convergence and to reach a global minimum. This paper introduces a two-stage approach for failure identification in a steel beam. In the first step, the presence of defects and their positions are identified by modal indices. In the second step, a feedforward neural network, improved by a hybrid particle swarm optimization and gravitational search algorithm, namely FNN-PSOGSA, is used to quantify the severity of damage. Finite element (FE) models of the beam for two damage scenarios are used to certify the accuracy and reliability of the proposed method. For comparison, a traditional ANN is also used to estimate the severity of the damage. The obtained results prove that the proposed approach can be used effectively for damage detection and quantification.

ispartof: COMPOSITE STRUCTURES vol:287 status: published

Vibration-based damage identification is a non-destructive method that enables the health monitoring of civil infrastructures. It aims to detect the presence and location of damage by measuring changes in the vibration characteristics of these structures. Unfortunately, the most popular vibrational parameters – natural frequencies and modal displacements – have a low sensitivity to certain types of local damage. Modal strains and curvatures, on the other hand, can be sufficiently sensitive to local damage, but monitoring modal strains is challenging. Indeed, the strain amplitudes can be in the sub-microstrain range (

Damage can be detected by vibration responses of a structure. Damage changes the modal properties such as natural frequencies, mode shapes, and damping ratios. Natural frequency is one of the most frequently used damage indicators. In this paper, the natural frequency is used to monitor damage in a free-free beam. The modal properties of the intact free-free beam are identified based on a setup of 15 accelerometers. A finite element model is used to model the free-free beam. Three models are considered: beam (1D), shell (2D), and solid (3D). The numerical models are updated based on the first five bending natural frequencies. The free-free beam is damaged by a rectangle cut. The experiment is re-setup and the model properties of the damaged beam are re-identified. The cuttings are modeled in the numerical simulations. The first five numerical bending natural frequencies of the damaged beam are compared with the experimental ones. The results showed that the 1D beam element model has the highest errors, while the 2D and 3D models have approximately the same results. Therefore, the 2D representation can be used to model the damaged beam for fast computation.

© 2018 Elsevier Ltd Vibration-based damage identification has been widely studied in the field of structural health monitoring (SHM) for several decades. It is well known, however, that low-order modal parameters, being among the most frequently used, are not sensitive to local damage. A suitable methodology is therefore needed to extract such damage features from the dynamic response of structures. In the present work, local bending behavior of cables is studied for damage identification. First, the dynamic response of a cable is decomposed into evanescent wave and propagating wave components. It is proven that the contribution of the evanescent wave is spatially concentrated, and is sensitive to local damage. A signal transform is proposed next, which allows the estimation of the wave components from the measured cable response. The reflection coefficient of the evanescent wave (REW), which can be calculated from the estimated wave coefficients, depends only on the characteristics of the local discontinuity, and proves to be a robust indicator for local damage. The feasibility of the proposed methodology is studied by means of a simulated experiment, considering a cable model with two locally damaged parts. The results show that the intensity of REW is significantly higher near the damage locations, allowing damage localization. From the estimated REW near the damage locations, the damage levels can be estimated, showing the potential of this methodology for damage assessment of cable structures. ispartof: JOURNAL OF SOUND AND VIBRATION vol:442 pages:527-551 status: published

Railway bridges are susceptible to the problems of fracture and fatigue because they endure millions of stress cycles under moving train load during their service life. This may lead to stiffness reduction in truss joints of the truss bridges and reduce the operational effectiveness of the bridge. In this paper, a hybrid algorithm based on Artificial Neural Network (ANN) coupled with an evolutionary algorithm, i.e. Genetic Algorithm (GA) is proposed to analyze the stiffness reduction of truss joints in a railway bridge. GA is employed to determine training parameters and overcome the local minimum problems of ANN. Natural frequencies are selected as input data, whereas output data is damage characteristics (locations and levels) of truss joints. The results demonstrate that ANN-GA determines vibration behavior and damages of the considered structure accurately, comprising single stiffness reduction of truss joints as well as multiple stiffness reduction scenarios.

Vibration-based model updating is widely accepted as an effective technique to construct a baseline model for structural health monitoring. In this paper, the effects of half-open cross-section and orthotropic deck on the global vibration behaviour of a truss bridge are considered. To do that, an FE model using ANSYS is built with variations of half-open cross-section area of bars and the stiffness of the orthotropic steel deck. The cross-section area of bars are assigned a reduced coefficient, less than and equal to one. Based on a vibration test, a hybrid optimization algorithm, known as Hybrid Particle Swarm Optimization and Gravitational Search Algorithm (PSOGSA), is applied to reduce the differences in frequencies and mode shapes between measurements and FE simulations. Finally, the effective coefficients of the cross-section as well as the stiffness of steel deck are obtained from the updated model.

Recent advances in optical sensing and signal processing technologies have resulted in strain sensing systems that can accurately capture the very small dynamic strain levels occurring in civil structures during ambient excitation. In particular, it has been demonstrated in a laboratory environment that the combination of fiber Bragg grating (FBG) sensors with an acquisition system that interrogates those sensors with a swept monochromatic signal, can accurately capture RMS strain values with an order of magnitude of 0.01 microstrain. In the present work, the potential of this novel sensing technology for civil structural health monitoring is explored for a steel bow-string railway bridge that has been monitored for over one year. Both sides of the bridge deck have been instrumented with chains of multiplexed FBG sensors that are fixed to the bridge at discrete points. In this way, dynamic macro-strains are measured in a continuous way along the entire length of the bridge. During the monitoring period, the bridge underwent a retrofitting operation, in which all connections between the hangers and the bridge’s deck and arch were altered. The influence of this retrofit on the monitored eigenfrequencies and strain mode shapes has been investigated and compared with their regular environmental variation. As the retrofitting resulted in global mass and stiffness changes of the structure, it is found to have an influence mainly on the eigenfrequencies, except when it induces an interaction between previously well-separated modes: in that case the strain mode shapes are also affected.

Vibration monitoring from strain data is a promising alternative to acceleration-based monitoring because a dense measurement grid can be achieved at a relatively low cost and because strain mode shapes are more sensitive to local stiffness changes than displacement mode shapes. However, the feasibility of monitoring strain mode shapes of full-scale civil structures, where the operational dynamic strain levels are of very low amplitude and temperature changes can influence the modal characteristics, has remained an open question. The present work provides a proof of concept in which the deck of a steel bowstring railway bridge is instrumented with 80 Fiber-optic Bragg Grating strain sensors, multiplexed in four fibers, that are interrogated with a technique that achieves high accuracy and precision. For more than a year, the natural frequencies and strain mode shapes of 10 modes have been automatically identified from operational strain time histories, with typical root- mean-square values of 0.01 microstrain, on an hourly basis. Furthermore, using these modal data, the influence of temperature fluctuations and that of a retrofitting of the hangers connecting the bridge deck and the bow, which took place during the monitoring period, are extensively investigated. Both have an influence on the overall stiffness of the bridge and therefore they result in clear changes in the natural frequencies. They do not have an influence on the local stiffness and therefore they do not influence the strain mode shapes, except when the retrofitting induces an interaction between previously well-separated modes.

Over the past decades European societies have become increasingly diverse. This diversity in culture, education, and language significantly impacts neuropsychological assessment. Although several initiatives are under way to overcome these barriers - e.g. newly developed and validated test batteries - there is a need for more collaboration in the development and implementation of neuropsychological tests, such as in the domains of social cognition and language. To address these gaps in cross-cultural neuropsychological assessment in Europe, the European Consortium on Cross-Cultural Neuropsychology (ECCroN) was established in 2019. ECCroN recommends taking a broad range of variables into account, such as linguistic factors, literacy, education, migration history, acculturation and other cultural factors. We advocate against race-based norms as a solution to the challenging interpretation of group differences on neuropsychological tests, and instead support the development, validation, and standardization of more widely applicable/cross-culturally applicable tests that take into account interindividual variability. Last, ECCroN advocates for an improvement in the clinical training of neuropsychologists in culturally sensitive neuropsychological assessment, and the development and implementation of guidelines for interpreter-mediated neuropsychological assessment in diverse populations in Europe. ECCroN may impact research and clinical practice by contributing to existing theoretical frameworks and by improving the assessment of diverse individuals across Europe through collaborations on test development, collection of normative data, cross-cultural clinical training, and interpreter-mediated assessment. Accepted for publication ispartof: Computer Applications In Engineering Education vol:36 issue:3 pages:1-18 ispartof: location:England status: published

This paper proposes a hybrid optimization algorithm combining particle swarm optimization (PSO) with genetic algorithm (GA) to update a railway bridge. PSO is an evolutionary optimization algorithm based on global search techniques to look for the best solution. Nevertheless, since PSO relies crucially on the quality of initial particles, it may reduce its effectiveness and robustness in tacking optimization issues. If the positions of initial populations are too far from the global best, it is challenging to determine the best solution. To overcome these shortcomings, we propose a hybrid optimization algorithm applying the advantages of both PSO and GA. GA is first used to generate the most elite populations based on its crossover and mutation characteristics. Those populations are then employed to seek the best solution based on the global search capacity of PSO. The experimental measurements of the railway bridge are carried out under ambient vibrations used to validate the proposed algorithm (PSO-GA). The result demonstrates that PSO-GA, GA, and PSO possibly determine uncertain parameters of the bridge exactly and PSO-GA surpasses GA alone and PSO alone in terms of convergence level and accuracy.

Structural Health Monitoring is an important field that involves the continuous measuring of the structural status of infrastructures. In order to be able to detect the damage status, data collected from sensors have to be processed to identify the difference between the damaged and the undamaged states. There exist machine learning techniques attempting to extract features from vibration data, however, they require prior knowledge about the factors affecting the structure. In this paper, we propose a novel method of damage detection using a convolution neural network and a handcrafted feature extraction. This method uses a convolution neural network to extract deep features in time series and uses handcrafted features to find a statistically significant correlation of each lagged features in time series data. These two types of features are combined to increase discrimination ability compared to deep features only. Finally, the neural network will be used to classify the time series into normal and damaged states. The accuracy of damaged detection was tested on a benchmark dataset from Los Alamos National Laboratory and the result shows that hybrid features provided a highly accurate damage identification.

Vibration-based monitoring was performed on a short-span skewed highway bridge on the basis of wireless measurements. By means of operational modal analysis, highly accurate modal results (frequencies and mode shapes) were extracted by using a self-developed wireless acquisition system, for which the performance was verified in the field. In order to reproduce the experimental modal characteristics, a refined finite element model was manually tuned to reduce the idealization errors and then updated with the sensitivity method to reduce the parametric errors. It was found that to build a reliable Finite element (FE) model for application in structural health monitoring, the effects of superelevation and boundary conditions of a skewed bridge should be taken into account carefully. Open Access ispartof: Applied Sciences-Basel vol:10 issue:7 pages:1-13 status: published

Mode shape-based method has shown its dominance in failure identification of beams. However, it is a challenge for fault detection in a plate structure. It requires combined information in two directions to determine the damage location. In this study, a damage index, namely mode shape derivative based damage identification (MSDBDI), is applied to localize damage in fixed-free plate structures. Two-dimensional (2D) displacement mode shapes and their derivatives are used to identify the MSDBDI index. It is a fact that this indicator is sometimes stuck in localizing damage, and it cannot indicate the damage severity. Hence, a coupled model between an artificial neural network (ANN) and antlion optimizer (ALO), so-called ALOANN is used to overcome this drawback. In this method, ALO instead of a backpropagation algorithm is used to look for the best initial values of learnable parameters of ANN i.e. weights and biases through mean squared error (MSE). These obtained parameters are added to ANN for damage identification. The efficiency of the proposed approach is tested with two numerical studies of the plate structures with single and multiple damage scenarios. In the first application, damage scenarios in a intact plate are detected. In the second application, ALO first is used to build an FE model of a composite structure based on a vibration experiment. Then the slab of the updated model is assumed to be suffered several damage scenarios. In both applications, failures are localized by using damage index. Then, the proposed approach is used to quantify the corresponding extent by means of changes in frequencies and displacement mode shapes. Values of damage index are achieved from modal properties of one or three out of the first five modes of the two considered structures. A conventional ANN also is investigated for comparison. Results of damage identification indicate that the damage indicator coupled with ALOANN show better performance in localization and quantification compared with using ANN alone even when a noise level is assigned to modal properties.

Vibration-based Structural Health Monitoring is a non-destructive condition assessment method that exploits damage-related changes in the dynamic characteristics of a structure. The sensitivity of some commonly employed dynamic characteristics such as natural frequencies, to local damage of moderate severity might be lower than their sensitivity to environmental influences such as temperature. Modal strains are more sensitive to local damage, but the accurate dynamic monitoring of the very low strain levels that occur in civil engineering structures under ambient excitation became possible only recently. In this work, the influence of both temperature and damage on modal strains is experimentally investigated for a prestressed concreted beam in controlled laboratory conditions. Not only the modal strains themselves are considered, but also the neutral axis position under bending deformation, which relates directly to the bending stiffness. It is found that the induced temperature changes in the concrete beam do not have a measurable influence on the modal strains and neutral axis positions. The cracks that are induced in the beam by external loading in a progressive damage test do have a clear and local influence on the strain mode shapes and neutral axis positions, even at low damage levels.

Vibration-based methods are among the most widely studied in structural health monitoring (SHM). It is well known, however, that the low-order modes, characterizing the global dynamic behaviour of structures, are relatively insensitive to local damage. Such local damage may be easier to detect by methods based on wave propagation which involve local high frequency behaviour. The present work considers the numerical analysis of wave propagation in cables. A two-step approach is proposed which allows taking into account the cable sag and the distribution of the axial forces in the wave propagation analysis. In the first step, the static deformation and internal forces are obtained by the finite element method (FEM), taking into account geometric nonlinear effects. In the second step, the results from the static analysis are used to define the initial state of the dynamic analysis which is performed by means of the spectral element method (SEM). The use of the SEM in the second step of the analysis allows for a significant reduction in computational costs as compared to a FE analysis. This methodology is first verified by means of a full FE analysis for a single stretched cable. Next, simulations are made to study the effects of damage in a single stretched cable and a cable-supported truss. The results of the simulations show how damage significantly affects the high frequency response, confirming the potential of wave propagation based methods for SHM.

© 2017 Elsevier Ltd This paper aims at assessing the influence of environmental parameters on the modal characteristics of age–old masonry constructions. The results of a long–term ambient vibration monitoring of the San Frediano bell tower in Lucca (Italy) are reported. The tower, dating back to the 11th century, has been fitted along its height with four triaxial seismometric stations, which were left active for about one year. Data from the monitoring system have been processed via the Stochastic Subspace Identification Method in order to identify the tower's modal characteristics and their variations over the year. The dependence of the tower's frequencies on the ambient temperature was first studied and simulated via simple auto–regressive models. Then, some output–only models based on the principal component analysis (PCA) were applied, under the hypotheses of both linear and nonlinear (Kernel PCA) dependence of the natural frequencies on the unknown environmental parameters. The results indicate PCA to be an effective tool for detecting changes in the dynamic characteristics of masonry constructions. ispartof: Engineering Structures vol:156 pages:175-187 status: published

Vibration monitoring from strain data is a promising alternative to the more conventional acceleration-based monitoring because a dense measurement grid can be achieved at a relatively low cost and because strain mode shapes are more sensitive to local stiffness changes than displacement mode shapes. However, the feasibility of monitoring strain mode shapes of full-scale civil structures, where the operational dynamic strain levels are of very low amplitude and temperature changes can influence the modal characteristics, has remained an open question. The present work provides a proof of concept in which the deck of a steel tied arch railway bridge is instrumented with eighty Fiber-optic Bragg Grating strain sensors, multiplexed in four fibers, that are interrogated with a technique that achieves high accuracy and precision. For more than a year, the natural frequencies and strain mode shapes of ten modes have been automatically identified from operational strain time histories, with typical root-mean-square values of 0.01 microstrain, on an hourly basis. Furthermore, using these modal data, the influence of temperature fluctuations and that of a retrofitting of the hangers connecting the bridge deck and the arches, which took place during the monitoring period, are extensively investigated. Both have an influence on the overall stiffness of the bridge and therefore they result in clear changes in the natural frequencies. They do not have an influence on the local stiffness and therefore they do not influence the strain mode shapes, except when the retrofitting induces an interaction between previously well-separated modes.

Vibration-based structural health monitoring of civil structures relies on the repeated identification of dynamic structural characteristics of the structure from output-only vibration data. Natural frequencies and displacement mode shapes are the most commonly employed dynamic characteristics; yet their sensitivity to local damage of moderate severity is rather low with respect to their sensitivity to other factors such as temperature, necessitating data normalization. Strain mode shapes offer a higher sensitivity to local damage, but their accurate identification in a dense grid is challenging given the very small dynamic strain levels that are encountered under ambient excitation. In this article, a method is presented for tackling this challenge. It consists of three stages. First, fiber-optic Bragg grating strain sensors are attached to the structure and interrogated with a tunable laser performing a wavelength sweep. In this way, the measured strain amplitudes have the required accuracy but synchronization errors are introduced between the different Bragg sensors. Second, a modal analysis is performed on the dynamic strain data using an accurate parametric system identification technique. This is followed by a synchronization step which compensates for the delays introduced by the wavelength sweep. Finally, the synchronized strain mode shapes are employed as damage-sensitive features, either directly or via a newly proposed quantity, the top-to-bottom strain ratio. The method is validated by progressive damage testing of a complex, prestressed concrete "roof" beam, reinforced with steel fibers. It is observed that the proposed method can identify both the presence and the location of the damage in a relatively early stage. ispartof: Structural Health Monitoring vol:17 issue:6 pages:1441-1459 status: published

Quasi-periodic structures, as found in multi-span bridges, multi-bay and multi-storey buildings, are often characterized by a high modal density, even at low frequencies. This clustering of modes poses challenges in operational modal analysis, as well as in finite element model updating, where a pairing of experimental and numerically predicted modes is required. To overcome these difficulties, an alternative method for model updating based on the so-called free wave characteristics has recently been proposed. In this work, this method is applied to calibrate a finite element model of the K032 viaduct of the A11 highway in Bruges, Belgium. Vibration measurements are conducted to experimentally determine the free wave characteristics of the viaduct. In the model updating, the discrepancy between the calculated and identified free wave characteristics is minimized. The results are validated by comparing measured frequency response functions with those obtained from the updated finite element model.

Finite element (FE) based structural health monitoring (SHM) algorithms seek to update structural damage indices through solving an optimisation problem in which the difference between the response of the real structure and a corresponding FE model to some excitation force is minimised. These techniques, therefore, exploit advanced optimisation algorithms to alleviate errors stemming from the lack of information or the use of highly noisy measured responses. This study proposes an effective approach for damage detection by using a recently developed novel swarm intelligence algorithm, i.e. the marine predator algorithm (MPA). In the proposed approach, optimal foraging strategy and marine memory are employed to improve the learning ability of feedforward neural networks. After training, the hybrid feedforward neural networks and marine predator algorithm, MPAFNN, produces the best combination of connection weights and biases. These weights and biases then are re-input to the networks for prediction. Firstly, the classification capability of the proposed algorithm is investigated in comparison with some well-known optimization algorithms such as particle swarm optimization (PSO), gravitational search algorithm (GSA), hybrid particle swarm optimization-gravitational search algorithm (PSOGSA), and grey wolf optimizer (GWO) via four classification benchmark problems. The superior and stable performance of MPAFNN proves its effectiveness. Then, the proposed method is applied for damage identification of three numerical models, i.e. a simply supported beam, a two-span continuous beam, and a laboratory free-free beam by using modal flexibility indices. The obtained results reveal the feasibility of the proposed approach in damage identification not only for different structures with single damage and multiple damage, but also considering noise effect.

The presence of damage in a civil structure alters its stiffness and consequently its modal characteristics. The identification of these changes can provide engineers with useful information about the condition of a structure and constitutes the basic principle of the vibration-based structural health monitoring. While eigenfrequencies and mode shapes are the most commonly monitored modal characteristics, their sensitivity to structural damage may be low relative to their sensitivity to environmental influences. Modal strains or curvatures could offer an attractive alternative but current measurement techniques encounter difficulties in capturing the very small strain (sub-microstrain) levels occurring during ambient, or operational excitation, with sufficient accuracy. This paper investigates the ability to obtain sub-microstrain accuracy with standard fiber-optic Bragg gratings using a novel optical signal processing algorithm that identifies the wavelength shift with high accuracy and precision. The novel technique is validated in an extensive experimental modal analysis test on a steel I-beam which is instrumented with FBG sensors at its top and bottom flange. The raw wavelength FBG data are processed into strain values using both a novel correlation-based processing technique and a conventional peak tracking technique. Subsequently, the strain time series are used for identifying the beam's modal characteristics. Finally, the accuracy of both algorithms in identification of modal characteristics is extensively investigated.

The paper focuses on damage detection of civil engineering structures and especially on concrete bridges. A method for structural health monitoring based on vibrational measurements is presented and discussed. Experimentally identified modal parameters (eigenfrequencies, mode shapes and modal masses) of bridge structures are used to calculate the inverse stiffness matrix, the so-called flexibility matrix. By monitoring of the stiffness matrix, damage can easily be detected, quantified and localized by tracking changes of its individual elements. However, based on dynamic field measurements, the acquisition of the flexibility matrix instead of the stiffness matrix is often the only choice and hence more relevant for practice. But the flexibility-based quantification and localisation of damage are often possible but more difficult, as it depends on the type of support and the location of the damage. These issues are discussed and synthetized, that is an originality of this paper and is believed useful for engineers in the damage detection of different bridge structures. First the theoretical background is briefly repeated prior to the illustration of the differences between stiffness and flexibility matrix on analytical and numerical examples. Then the flexibility-based detection is demonstrated on two true bridges with real-time measurement data and the results are promising.

Modal testing of large structures is usually performed in multiple setups in order to obtain sufficiently detailed information on the mode shapes with a limited number of sensors. Several reference sensors are kept fixed during the entire measurement, while the other so-called roving sensors are moved in different setups. A careful choice of the reference and roving sensor positions is necessary for a successful modal analysis. An optimal sensor placement strategy which was developed for multi-setup modal identification is applied to the modal testing of a nearly repetitive office building. Based on the strategy the optimal reference sensor locations are designed for the building. As a comparison, badly chosen reference sensor locations are also studied. The optimal sensor placement strategy is verified by comparing the modal identification results obtained from the optimally selected reference sensor locations to those obtained from the badly chosen reference sensor locations. The experimental results show that the optimal sensor placement strategy for multi-setup modal testing allows extracting more information from the measurement. Two roving sensor strategies, i.e. a cluster and a uniform distribution of roving sensors, are also investigated. It is found that a uniform distribution of roving sensors is preferred over a cluster of roving sensors.

A vibration serviceability assessment of footbridges is required in design stage to evaluate the response to human-induced excitation. If the calculated response does not meet the desired criteria for vibration comfort, a Tuned Mass Damper (TMD) can be installed as a vibration mitigation device. The TMD mass, stiffness and damping constant must be tuned to the modal parameters of the structure to obtain the desired response reduction. The response prediction and TMD design rely on a good knowledge of the modal parameters of the footbridge and usually assume that the response is governed by a perfect harmonic loading causing resonance. Differences between the predicted and actual modal parameters may lead to both an unreliable response prediction and a suboptimal TMD performance. Therefore, a robust design of the TMD is proposed accounting for uncertainties in the modal parameters. Realistic walking scenarios of continuous pedestrian traffic are simulated to obtain an effective response reduction by the TMD. In this contribution, the robust design of a TMD is demonstrated for a slender footbridge accounting for reasonable levels of uncertainties in the modal parameters. Numerical optimisation is therefore applied and vibration serviceability requirements are imposed as design constraints. The obtained TMD parameters are compared to those obtained from the formulae proposed by Asami, determining the stiffness and damping constant as a function of its mass. It is found that the TMD mass can be further reduced when the TMD parameters are tuned independently from each other compared to the classical design according to Asami but a higher computation cost is needed. Furthermore, an increasing degree of robustness against variations in the modal parameters is obtained by increasing the TMD mass and damping.

© 2017 The Authors. Published by Elsevier Ltd. For the design of slender footbridges, the vibration serviceability under pedestrian excitation is often the governing criterion. In design stage, vibration levels are predicted using simplified load models that are extrapolated from single-person force models to represent the effect of a crowd. However, these load models disregard Human-Human Interaction (HHI) and Human-Structure Interaction (HSI). This contribution investigates the effect of human-human interaction on the resulting structural response. A social force model is applied to simulate the realistic pedestrian traffic. The time-varying position and velocity of each pedestrian in the crowd are transferred into the necessary inputs for detailed step-by-step simulations of the pedestrian-induced forces and the resulting pedestrian-induced vibrations. The results show that accounting for HHI results into a reduced global walking speed. Also, the inter-person variability of the step frequencies is lower than when HHI is disregarded. As an alternative to the computationally expensive social force model, the effect of HHI is translated into an equivalent distribution of step frequencies of the pedestrians in the crowd. The results show that this simplified model allows for a very good approximation of the HHI effects on the resulting crowd-induced loading and structural response. ispartof: pages:2907-2912 ispartof: Procedia Engineering vol:199 pages:2907-2912 ispartof: X International Conference on Structural Dynamics location:Rome, Italy date:10 Sep - 13 Sep 2017 status: published

This paper presents an approach for stiffness identification of node joints of a large-scale truss bridge (Nam O Bridge in Vietnam) based on vibration measurements and model updating. Vibrations are recorded under ambient conditions using piezoelectric sensors. Excitation is due to wind, micro-tremors, or train passage. From these vibrations, modal parameters are extracted. Modal analysis is also performed using a finite element model created in MATLAB. Afterwards, model updating is applied to minimize the discrepancy between numerical and experimental modal analysis results. Evolutionary algorithm such as particle swarm optimization (PSO) based on a global search technique is employed. Three scenarios of boundary conditions of truss joints (pin, rigid, and semi-rigid) are considered. The result of model updating shows that semi-rigid (using rotational springs) joint conditions represents correctly the dynamic behavior of the considered bridge.

Nowadays, many methods have been employed to estimate the cable tension in a cable-stayed bridge. In these methods, the non-destructive test has a widespread application thanks to its feasibility and effectiveness. The vibration-based approach determines the cable tension based on natural frequency. In fact, there are many factors that can influence the tension values. Some of them are derived from objective factors, others from subjective factors. In this paper, a subjective issue, so-called measurement technique, is studied. A cable, without fillings between the HDPE duct and strands, is the main object of this study. In the first step, some scenarios related to the placement of sensors are conducted. The former considers the tension between different positions of sensors, directly and indirectly contact conditions with strands, under the ambient and pulling-rope excitation. The latter takes into account the directions, out-of-plane and in-plane, of sensor placement on the strands or HPDE duct. In the next step, the relationship between the force and frequency is utilized to identify the mean tension of the cable. In the finally step, results of cable tension estimation are compared with the real-time monitoring and finite element (FE) simulation. From the comparison, the effects of measuring technique in situ are discussed in the conclusion.

Vibration-based Structural Identification has become an increasingly popular experimental technique, e.g. for model calibration, control of structural behavior during construction, assessment of the efficiency of structural repair, … Main breakthrough was the development of operational modal analysis (OMA), avoiding the use of artificial vibration sources. Vibration-based Structural Health Monitoring is based on the principle that modal parameters of a structure are stiffness dependent. Changes in natural frequencies, damping ratios, mode shapes or combinations can therefore be used as features to detect and to identify damage. Non model-based methods try to identify damage from a comparison of pure measurements (or features directly derived from them) before and after a seismic event without relying on a (numerical) model. In most cases only the first (alarm) level of damage can be reached. The advantage of model-based methods is that they can go further in the damage identification process. They intend to localize and to quantify the damage. Disadvantage is the need to construct a reliable model. Mostly, linear models are used representing the quasi-linear state at small ambient vibrations before and after an earthquake, although there is an interest to use non-linear models for simulating the structural behavior during the earthquake.

Footbridges are often designed as slender structures, sensitive to human-induced excitation. In the case where the prevailing vibration serviceability requirements are not met, vibration reduction measures such as tuned mass dampers (TMDs) are needed. Both the prediction of the structural response and the design of the TMD rely on the modal parameters of the footbridge. Measurements after construction of footbridges have shown that it is difficult to accurately predict these modal parameters even with detailed finite element models based on structural drawings. Moreover, these parameters evolve in time due to changing environmental conditions and degradation. It is important to take into account these uncertainties in the vibration serviceability assessment as well as in the design of vibration reduction measures. The present paper proposes a robust optimisation approach for the design of a TMD which accounts for uncertainties in the modal parameters. The aim is to minimise the mass of the TMD, assumed as a measure for the cost, by tuning the mass, stiffness and damping values, while guaranteeing that vibration serviceability is satisfied for a range of possible values of the natural frequency and modal damping. In order to investigate the trade-off between the mass of the TMD (cost) and the level of uncertainty, a multi-interval approach is adopted. The optimal parameters are found to change significantly with the level of uncertainty. The TMD mass and damping increase for a higher level of uncertainty to satisfy the vibration serviceability requirements in all possible cases. publisher: Elsevier articletitle: Robust design of a TMD for the vibration serviceability of a footbridge journaltitle: Engineering Structures articlelink: http://dx.doi.org/10.1016/j.engstruct.2016.05.028 content_type: article copyright: © 2016 Elsevier Ltd. All rights reserved. ispartof: Engineering Structures vol:123 pages:408-418 status: published

This paper presents the detailed experimental results of 23 full-scale rectangular and I-shaped prestressed and reinforced concrete beams. All beams were designed to fail in shear and were subjecte...

The Bears’ Cage footbridge is a slender steel structure with a single span. Its dynamic behaviour is predicted based on a refined finite-element (FE) model and the vibration serviceability is assessed according to the current codes of practice. The assessment indicates a high susceptibility to human-induced vibrations with disturbing vibration levels even for sparse pedestrian densities. To validate the predicted behaviour of the structure, an extensive experimental study is performed including static deflection and dynamic vibration tests. The analysis shows that statically, the longitudinal movement of the supports on one side of the span can be considered unconstrained, indicating a behaviour of the sliding pot bearings as designed. Due to the footbridge’s arch-like shape, the longitudinal stiffness of the supports highly influences the natural frequency of the fundamental bending mode. The analysis shows that the longitudinal stiffness of sliding pot bearings and the structural inherent dampin...

This paper addresses a damage detection method based on changes in modal curvature combined with Convolutional Neural Network (CNN). The use of modal curvature for damage detection is very well known. Some methods require modal data of healthy structures as a reference, others do not. Gapped Smoothing Method (GSM) is a vibration-based damage detection methodology, which makes use of modal curvature for identifying the location of structural damage and does not require information of the intact structure. The Bo Nghi bridge is used as an illustrative example. This bridge consists of four T-shaped concrete simply supported girders. One single beam with the same length and cross-section of the bridge girder is modeled and used to extract numerical data to train the CNN. A CNN is trained by using images from the damage index of the GSM to classify the damage location in the numerical beam model. Finally, the finite element model of the bridge is built and used to model damage scenarios to test the trained CNN. The results indicate that the combination of GSM and CNN can be used for damage detection and localization.

Vibration-based force identification of cables has been studied for several decades. Most of this work relies on the natural frequencies of the cable for an estimation of the cable force. However, these natural frequencies are also affected by bending stiffness, sag effect and boundary conditions. In the present work, a two-step methodology is developed that allows taking into consideration these effects in the force identification. First, a segment of the cable is considered which is sufficiently short for the sag effect to be negligible. The axial force in this segment is estimated by fitting the measured response to the analytical solution for the transverse motion of the cable in the frequency domain. In this procedure, the bending stiffness is updated exploiting the fact that the estimated axial force should not depend on the frequency, while the boundary conditions do not need to be known. Next, an analytical solution of the static state of the entire cable is derived, taking into account the sag effect, bending stiffness and boundary conditions. The parameters of the entire cable model can then be updated, using the estimated value of the axis force at the location of the segment. Finally, the updated analytical model of the entire cable allows evaluating internal forces such as the cable force and bending moment, as required for estimating the stresses in the cable considering bending deformation. The feasibility of the proposed methodology is verified by means of numerical simulations considering measurement noise and an inaccurate initial guess of the bending stiffness, proving its potential for the health monitoring of cable structures.

The first two steps in Structural Health Monitoring (SHM) are to detect and localize damage in a structure. Transmissibility uses only output signals to extract the structural dynamic parameter and very sensitive to damage. In this paper, a simply supported beam is analyzed using a transmissibility-based damage detection methodology combined with Auto-Associative Neural Network (AANN). Firstly, the transmissibility is calculated between two points in the examined beam. Then a transmissibility indicator is taken as input for the neural network, which accounts only for the response data and is sensitive to damage. The target of AANN is the location and the severity of the damage. It is found that the network, after being trained, show good results and can be used to detect and localize damage.

© 2018 Elsevier Ltd The present paper proposes a strategy for the robust vibration serviceability assessment of footbridges. The response of the footbridge to vertical walking loading is predicted by simulating continuous flows of pedestrians on the structure. In the calculations, both variability in the walking characteristics and human-structure interaction effects are accounted for. Uncertainty in the modal parameters of the footbridge is considered in a multi-interval approach. The resulting range in predicted response levels is subsequently compared to vibration comfort criteria. A case study serves as illustration of the procedure. First, the response of the footbridge is evaluated in a multi-interval assessment showing how the predicted vibration levels are affected by uncertainties. The multi-interval approach allows understanding how uncertainties in the modal parameters of the structure affects the response. Second, the procedure is applied for the design of Tuned Mass Dampers (TMD). The TMD parameters are tuned by solving an optimisation problem such that an effective reduction of the accelerations under realistic walking scenarios is ensured. The total TMD mass is minimised, considered as the determining factor for the cost of the TMD. A higher TMD mass is needed to satisfy the vibration serviceability constraints for higher levels of uncertainty, showing the trade-off between cost and robustness. ispartof: Engineering Structures vol:171 pages:236-246 status: published

Multi-setup operational modal testing has been performed on a multi-span viaduct in Bruges, Belgium. The viaduct consists of two curved integral bridges with a length of about 800 m in parallel with each other. Each bridge has 23 spans and consists of nearly periodic parts. During the measurement, a number of reference sensors is kept fixed, while the other so-called roving sensors are moved in different setups. Modal identification is first carried out for each setup separately. Next, the identified modal data of different setups are merged. It is observed from the stabilization diagrams that the modes are very closely spaced, resulting in a very high modal density, even under 2 Hz. This results in challenges in the modal identification. First, a high model order needs to be considered, leading to a high computational cost. Second, it is difficult to match the partial mode shapes obtained from different setups, due to the closely spaced natural frequencies. This problem is resolved by comparing both the natural frequencies and the mode shapes at the reference locations in the mode matching.

The identification of changes in the dynamic characteristics which are related to structural damage is the core principle of the vibration-based structural health monitoring (VBSHM). By addressing these changes, engineers can acquire valuable information about the condition of structures. Eigenfrequencies are influenced by environmental factors and this influence can be high enough to completely mask the effect of damage while they are also global structural characteristics and they cannot asses the location of the damage. A promising alternative is the monitoring of modal strains which offers the benefit of high sensitivity to local damage. However, current measurement techniques encounter difficulties in capturing the very small strain (sub-microstrain) levels that occur during ambient or operational excitation with sufficient accuracy and precision.

© 2017 IOP Publishing Ltd. Vibration-based damage identification is a well-known method to support health monitoring of civil engineering structures. Damage in such structures can be identified by measuring changes of the natural frequencies, damping factors or modal displacements of the structure. However, this approach suffers from the low sensitivity of these natural frequencies and modal displacements to certain types of damage. Modal strains and curvatures can be more sensitive to local damage, but direct monitoring of these quantities with sufficient spatial resolution is not possible with current measurement techniques due to the very small strain levels (sub-microstrain) caused by ambient or operational excitation. To deal with this issue, we propose a novel mechanical transducer equipped with an optical fiber Bragg grating (FBG) sensor that enhances the sensitivity to strain with a factor larger than 30. The principle of operation of the transducer exploits a symmetric cantilever structure that enlarges the strain experienced by the FBG sensor compared to the strain applied to the transducer itself. We carried out dynamic and static tests to verify the ability of the strain-amplifying transducers to measure small-amplitude strain levels and to evidence the potential for carrying out FBG based modal strain measurements on concrete civil engineering structures. ispartof: Smart Materials and Structures vol:26 issue:7 pages:1-10 status: published

© 2017 Elsevier Ltd Modal tests on large structures are often performed in multiple setups for practical reasons. Several sensors are kept fixed as reference sensors over all setups, while the other, so called roving sensors, are moved from one setup to another. This paper develops an optimal sensor placement strategy for multi-setup modal identification, which simultaneously optimizes the locations of the reference sensors and roving sensors. As an optimality criterion, the Information Entropy is adopted, which is a scalar measure of uncertainty in the Bayesian framework. The focus in the application goes to repetitive structures where modes typically occur in clusters, with closely spaced natural frequencies and similar wavelengths. The proposed strategy is illustrated for selecting optimal positions of uni-axial sensors for a repetitive frame structure. The influence of the number of reference sensors and two strategies for positioning roving sensors, i.e. a cluster and a uniform distribution of roving sensors, are investigated. The number of reference sensors is found to be preferably equal to or larger than the number of modes to be identified. In this case, the information content, as quantified by the Information Entropy, is not very sensitive to the roving sensor strategy. If less reference sensors are used, it is highly preferred to distribute the roving sensors uniformly over the structure instead of clustering them. The proposed strategy has been validated by an experimental modal test on a floor of an office building of KU Leuven, which has a nearly repetitive structural layout. The results show how optimally locating sensors allows extracting more information from the data. Though the focus is on applications involving repetitive structures, the proposed strategy can be applied to multi-setup modal identification of any large structure. ispartof: Journal of Sound and Vibration vol:401 pages:214-232 status: published