Search

Summary: This paper presents a benchmark problem for the structural health monitoring community to study tall buildings. The benchmark building is called the Green Building located at the Massachusetts Institute of Technology campus, with 21 stories above the ground (83.7 m) and a basement (3.8 m) connecting to the Massachusetts Institute of Technology tunnel system. This building was constructed as cast‐in‐place reinforced concrete and instrumented with 36 accelerometers to measure the building translational, torsional and vertical responses. The benchmark problem includes the detailed description of this building, 7 field measurement data sets (4 ambient data sets, 1 data set under an unidentified event, 1 data set under the excitation of fireworks, and 1 earthquake data set), and finite element models (both full‐scale and condensed models). The Green Building has an identifiable soil‐structure interaction behavior and the base rocking movement brings significant components into the building response. To decouple the rocking effect, storey measurement condensation and rocking response determination are discussed in this paper. A blind source separation approach is finally applied to identify the modal characteristics and quantify the rocking components. The benchmark data and models are open to the public for algorithmic development and validation. [ABSTRACT FROM AUTHOR]

This paper introduces a feature for detecting damage or changes in structures, the continuous symmetry measure, which can quantify the amount of a particular rotational, mirror, or translational symmetry in a mode shape of a structure. Many structures in the built environment have geometries that are either symmetric or almost symmetric, however damage typically occurs in a local manner causing asymmetric changes in the structure's geometry or material properties, and alters its mode shapes. The continuous symmetry measure can quantify these changes in symmetry as a novel indicator of damage for data-based structural health monitoring approaches. This paper describes the concept as a basis for detecting changes in mode shapes and detecting structural damage. Application of the method is demonstrated in various structures with different symmetrical properties: a pipe cross-section with a finite element model and experimental study, the NASA 8-bay truss model, and the simulated IASC-ASCE structural health monitoring benchmark structure. The applicability and limitations of the feature in applying it to structures of varying geometries is discussed. [ABSTRACT FROM AUTHOR]

In structural health monitoring of composite structures, one important task is to detect and identify the low-velocity impact events, which may cause invisible internal damages. This paper presents a novel approach for simultaneously identifying the impact location and reconstructing the impact force time history acting on a composite structure using dynamic measurements recorded by a sensor network. The proposed approach consists of two parts: (1) an inner loop to reconstruct the impact force time history and (2) an outer loop to search for the impact location. In the inner loop, a newly developed inverse analysis method with Bayesian inference regularization is employed to solve the ill-posed impact force reconstruction problem using a state-space model. In the outer loop, a nonlinear unscented Kalman filter (UKF) method is used to recursively estimate the impact location by minimizing the error between the measurements and the predicted responses. The newly proposed impact load identification approach is illustrated by numerical examples performed on a composite plate. Results have demonstrated the effectiveness and applicability of the proposed approach to impact load identification. Copyright © 2016 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

This paper investigates a new probabilistic strategy for Bayesian model updating using incomplete modal data. Direct mode matching between the measured and the predicted modal quantities is not required in the updating process, which is realized through model reduction. A Markov chain Monte Carlo technique with adaptive random-walk steps is proposed to draw the samples for model parameter uncertainty quantification. The iterated improved reduced system technique is employed to update the prediction error as well as to calculate the likelihood function in the sampling process. Since modal quantities are used in the model updating, modal identification is first carried out to extract the natural frequencies and mode shapes through the acceleration measurements of the structural system. The proposed algorithm is finally validated by both numerical and experimental examples: a 10-storey building with synthetic data and a 8-storey building with shaking table test data. Results illustrate that the proposed algorithm is effective and robust for parameter uncertainty quantification in probabilistic model updating of buildings. [ABSTRACT FROM AUTHOR]

Summary: Structural health monitoring (SHM) applications generally utilise high sampling rates, which low‐power wireless protocols are not well equipped to handle. Smart sensing approaches can overcome this, by using the processing capability of the sensor nodes to reduce the volume of data prior to communication. Most smart sensing approaches are preprogrammed and static. This causes two issues: First, the data processing logic cannot be easily modified, making it difficult to update and improve algorithms once deployed. Secondly, there is limited ability to adapt to changes in the environment or degradation of hardware. To address these problems, we have developed a system that allows users to remotely specify their computational logic on the fly in a MapReduce style syntax. We model these user‐specified tasks as a directed acyclic graph, and combine this model with statistics of the performance of each node in the network to formulate an optimisation problem. Solving this problem optimally allocates data processing operations to nodes in the network, such that the total time spent is minimised. We demonstrate a field deployment of this system, which illustrates the advantages of the proposed approach for a typical SHM application, and examines robustness of the system under environmental variations. [ABSTRACT FROM AUTHOR]

Summary: We present a computational methodology for structural identification and damage detection via linking the concepts of seismic interferometry and Bayesian inference. A deconvolution‐based seismic interferometry approach is employed to obtain the waveforms that represent the impulse response functions with respect to a reference excitation source. Using the deconvolved waveforms, we study the following two different damage detection methods that utilize shear wave velocity variations: the arrival picking method and the stretching method. We show that variations in the shear wave velocities can be used for qualitative damage detection and that velocity reduction is more evident for more severely damaged states. Second, a hierarchical Bayesian inference framework is used to update a finite element model by minimizing the gap between the predicted and the extracted time histories of the impulse response functions. Through comparison of the model parameter distributions of the damaged structure with the updated baseline model, we demonstrate that damage localization and quantification are possible. The performance of the proposed approach is verified through two shake table test structures. Results indicate that the proposed framework is promising for monitoring structural systems, which allows for noninvasive determination of structural parameters. [ABSTRACT FROM AUTHOR]

Summary: Computational modeling, in addition to data analytics, plays an important role in structural health monitoring (SHM). The high‐fidelity computational model based on the design and construction information provide important dynamics information of the structure and, more importantly, can be updated against field measurements for SHM purposes such as damage detection, response prediction, and reliability assessment. In this paper, we present a unique skyscraper (Al‐Hamra Tower) located in Kuwait City and its high‐fidelity computational model using ETABS for structural health monitoring applications. The tower is made of cast‐in‐place reinforced concrete with a core of shear walls and two curved shear walls running the height of the building (approximately 413 m with 86 floors in total). Interesting static and dynamic characteristics of the tower are described. System identification, interferometry‐based wave propagation analysis, and wave‐based damage detection are performed using synthetic data. This work mainly presents the phase of numerical investigations, which serves as a basis for correlating the field monitoring data to the model of the building in future work. [ABSTRACT FROM AUTHOR]

Through advances in sensor technology and development of camera-based measurement techniques, it has become affordable to obtain high spatial resolution data from structures. Although measured datasets become more informative by increasing the number of sensors, the spatial dependencies between sensor data are increased at the same time. Therefore, appropriate data analysis techniques are needed to handle the inference problem in presence of these dependencies. In this paper, we propose a novel approach that uses graphical models (GM) for considering the spatial dependencies between sensor measurements in dense sensor networks or arrays to improve damage localization accuracy in structural health monitoring (SHM) application. Because there are always unobserved damaged states in this application, the available information is insufficient for learning the GMs. To overcome this challenge, we propose an approximated model that uses the mutual information between sensor measurements to learn the GMs. The study is backed by experimental validation of the method on two test structures. The first is a three-story two-bay steel model structure that is instrumented by MEMS accelerometers. The second experimental setup consists of a plate structure and a video camera to measure the displacement field of the plate. Our results show that considering the spatial dependencies by the proposed algorithm can significantly improve damage localization accuracy. [ABSTRACT FROM AUTHOR]

• A new novelty detection classifier is proposed. • Dependence structure of random variables was considered as the discriminant feature. • Kernel dependence analysis was used to handle arbitrarily high dimensional problems. • No prior information is needed about the dependence structure of random variables. • Over 14% FP reduction compared to gradient boosting and SVM in an SHM problem. In this study, we propose a new approach for novelty detection that uses kernel dependence techniques for characterizing the statistical dependencies of random variables (RV) and use this characterization as a basis for making inference. Considering the statistical dependencies of the RVs in multivariate problems is an important challenge in novelty detection. Ignoring these dependencies, when they are strong, may result in inaccurate inference, usually in the form of high false positive rates. Previously studied methods, such as graphical models or conditional classifiers, mainly use density estimation techniques as their main learning element to characterize the dependencies of the relevant RVs. Therefore, they suffer from the curse of dimensionality which makes them unable to handle high-dimensional problems. The proposed method, however, avoids using density estimation methods, and rather, employs a kernel method, which is robust with respect to dimensionality, to encode the dependencies and hence, it can handle problems with arbitrarily high-dimensional data. Furthermore, the proposed method does not need any prior information about the dependence structure of the RVs; thus, it is applicable to general novelty detection problems with no simplifying assumption. To test the performance of the proposed method, we apply it to realistic application problems for analyzing sensor networks and compare the results to those obtained by peer methods. [ABSTRACT FROM AUTHOR]

The problem of automatic damage detection in civil structures is complex and requires a system that can interpret collected sensor data into meaningful information. We apply our recently developed switching Bayesian model for dependency analysis to the problems of damage detection and classification. The model relies on a state-space approach that accounts for noisy measurement processes and missing data, which also infers the statistical temporal dependency between measurement locations signifying the potential flow of information within the structure. A Gibbs sampling algorithm is used to simultaneously infer the latent states, parameters of the state dynamics, the dependence graph, and any changes in behavior. By employing a fully Bayesian approach, we are able to characterize uncertainty in these variables via their posterior distribution and provide probabilistic estimates of the occurrence of damage or a specific damage scenario. We also implement a single class classification method which is more realistic for most real world situations where training data for a damaged structure is not available. We demonstrate the methodology with experimental test data from a laboratory model structure and accelerometer data from a real world structure during different environmental and excitation conditions. [ABSTRACT FROM AUTHOR]