Structural Health Monitoring (SHM) based on the vibration of structures has been very attractive subject for researchers in different fields such as: civil, aeronautical and mechanical engineering. System Identification (SI) and Vibration based Damage Identification (VBDI) are two main parts of SHM. A full-scale seven-story reinforced concrete (RC) wall building has been tested during October 2005 and January 2006 by the University of California at San Diego (UCSD). The building was excited through four historical California ground motions. The RC wall experienced different levels of damage, progressively under increasing intensity of ground motions. At different levels of damage, the building was subjected to ambient vibration tests and low-amplitude White Gaussian Noise (WGN) base excitation. In this study, the response of the structure to ambient vibration tests is used to identify damage using VBDD method. The frequency domain decomposition method (FDD) is used to identify the modal parameters of the building. Damage changes the modal properties (frequency, mode shape and damping) by reducing the stiffness. Therefore, changes in the vibration characteristics of the structure can be used to identify location and severity of damage. A mode shape curvature-based method is used to detect and localize damage. Also a data-driven technique based on Neural Networks has been developed to identify the damage in the structure. The results show a close correlation with the structural damage observed in the experimental study.
Structural Health Monitoring (SHM) based on the vibration of structures has been very attractive topic for researchers in different fields such as: civil, aeronautical and mechanical engineering. The aim of this paper is to compare three most common modal identification techniques such as Frequency Domain Decomposition (FDD), Stochastic Subspace Identification (SSI) and Continuous Wavelet Transform (CWT) to find modal properties (such as natural frequency, mode shape and damping ratio) of three story book shelf steel structure which was built in Concordia University Lab. The modified Complex Morlet wavelet have been selected for wavelet in order to use asymptotic signal rather than real one with variable bandwidth and wavelet central frequency. So, CWT is able to detect instantaneous modulus and phase by use of local maxima ridge detection.
Proc. SPIE. 10170, Health Monitoring of Structural and Biological Systems 2017
KEYWORDS: Roads, Detection and tracking algorithms, Sensors, Diagnostics, Buildings, Data acquisition, Data processing, Signal processing, Structural health monitoring, Modal analysis, Damage detection, System identification, Bridges
Structural Health Monitoring (SHM) using dynamic characteristics of structures is crucial for early damage detection. Damage detection can be performed by capturing and assessing structural responses. Instrumented structures are monitored by analyzing the responses recorded by deployed sensors in the form of signals. Signal processing is an important tool for the processing of the collected data to diagnose anomalies in structural behavior. The vibration signature of the structure varies with damage. In order to attain effective damage detection, preservation of non-linear and non-stationary features of real structural responses is important. Decomposition of the signals into Intrinsic Mode Functions (IMF) by Empirical Mode Decomposition (EMD) and application of Hilbert-Huang Transform (HHT) addresses the time-varying instantaneous properties of the structural response. The energy distribution among different vibration modes of the intact and damaged structure depicted by Marginal Hilbert Spectrum (MHS) detects location and severity of the damage. The present work investigates damage detection analytically and experimentally by employing MHS. The testing of this methodology for different damage scenarios of a frame structure resulted in its accurate damage identification. The sensitivity of Hilbert Spectral Analysis (HSA) is assessed with varying frequencies and damage locations by means of calculating Damage Indices (DI) from the Hilbert spectrum curves of the undamaged and damaged structures.
Proc. SPIE. 10170, Health Monitoring of Structural and Biological Systems 2017
KEYWORDS: Roads, Data modeling, Interference (communication), Buildings, Signal processing, Civil engineering, System identification, Bridges, Neodymium, Signal analyzers, Stochastic processes, Prototyping, Correlation function
Dynamic study is important in order to design, repair and rehabilitation of structures. It has played an important role in the behavior characterization of structures; such as: bridges, dams, high rise buildings etc. There had been substantial development in this area over the last few decades, especially in the field of dynamic identification techniques of structural systems. Frequency Domain Decomposition (FDD) and Time Domain Decomposition are most commonly used methods to identify modal parameters; such as: natural frequency, modal damping and mode shape. The focus of the present research is to study the dynamic characteristics of typical timber masonry walls commonly used in Portugal. For that purpose, a multi-storey structural prototype of such wall has been tested on a seismic shake table at the National Laboratory for Civil Engineering, Portugal (LNEC). Signal processing has been performed of the output response, which is collected from the shaking table experiment of the prototype using accelerometers. In the present work signal processing of the output response, based on the input response has been done in two ways: FDD and Stochastic Subspace Identification (SSI). In order to estimate the values of the modal parameters, algorithms for FDD are formulated and parametric functions for the SSI are computed. Finally, estimated values from both the methods are compared to measure the accuracy of both the techniques.
Proc. SPIE. 9805, Health Monitoring of Structural and Biological Systems 2016
KEYWORDS: Mathematical modeling, Data mining, Data modeling, Data storage, Computer simulations, Clouds, Feature extraction, Data processing, Signal processing, Structural health monitoring, Damage detection, Machine learning, Feature selection, Double positive medium
Recently, data-driven models for Structural Health Monitoring (SHM) have been of great interest among many researchers. In data-driven models, the sensed data are processed to determine the structural performance and evaluate the damages of an instrumented structure without necessitating the mathematical modeling of the structure. A framework of data-driven models for online assessment of the condition of a structure has been developed here. The developed framework is intended for automated evaluation of the monitoring data and structural performance by the Internet technology and resources. The main challenges in developing such framework include: (a) utilizing the sensor measurements to estimate and localize the induced damage in a structure by means of signal processing and data mining techniques, and (b) optimizing the computing and storage resources with the aid of cloud services. The main focus in this paper is to demonstrate the efficiency of the proposed framework for real-time damage detection of a multi-story shear-building structure in two damage scenarios (change in mass and stiffness) in various locations. Several features are extracted from the sensed data by signal processing techniques and statistical methods. Machine learning algorithms are deployed to select damage-sensitive features as well as classifying the data to trace the anomaly in the response of the structure. Here, the cloud computing resources from Amazon Web Services (AWS) have been used to implement the proposed framework.
The Golden Boy statue was placed on top of the Manitoba Legislative Building in 1919 and has since served as a source of inspiration to all Manitobans. An inspection of this heritage structure was conducted in 2001and revealed that its steel supporting shaft had deteriorated significantly. A decision was made to take the statue down for restoration. A stronger, stainless steel shaft replaced the worn shaft and sensors including electrical strain gauges, accelerometers, fiber optic sensors and thermocouples were installed. Also, a web camera and wind meter were installed on the roof of the building. Data from the sensors and video feed from the web camera are available through the Internet to facilitate web-based Structural Health Monitoring (SHM) in real-time. The support shaft of the statue can be idealized as a single degree of freedom cantilever structure. Wind and acceleration data are used to estimate the strains experienced by the shaft near the base, which are then correlated with the actual strain recorded by the strain sensors. This correlation was difficult as the data contained various levels of measurement errors or noise, and the strain data must be isolated from the thermal strain. Finally, the observed strain was correlated with hourly peak wind velocities reported by Environment Canada and an empirical relationship between these quantities was established to obtain an estimate of the strain in the shaft based on wind velocity, which will detect malfunctions in the sensors or deterioration of the structure. The study provided an understanding of the statue's behavior under a range of wind speeds and confidence in the SHM system for continuous and long term monitoring. It established a baseline response and alternative methods for predicting the behavior of the statue that provide a foundation for comparison with future stages in the Structural Health Monitoring of the Golden Boy.