Mehrisadat Makki Alamdari

Spectral-based damage identification in structures under ambient vibration

AbstractThe motivation behind this paper is to develop a spectral-based damage detection and damage localization scheme using in-service ambient vibration in the context of non–model-based damage characterization. In this regard, a response parameter known as spectral moment is implemented for structural damage identification. The damage identification procedure starts with developing response power spectral density (PSD). The principal structural response features are then extracted from the frequency distribution of the spectrum using spectral moments. It is demonstrated that, although, spectral moment is a nonmodal characteristic of a process, it is related to modal parameters of a response signal since the spectral moment at each location is proportional to its corresponding modal vector. Hence, it is expected that due to damage occurrence spectral moment undergoes a variation. On this point, a damage sensitive feature is defined by comparing the spectral moments of two successive states of the struct...

Nonlinear Joint Model Updating in Assembled Structures

AbstractDynamic response of mechanical structures is significantly affected by joints. Joints introduce remarkable frictional damping and localized flexibility to the structure; hence, to obtain a more accurate representation of a systems dynamics, it is crucial to take these effects into account. This paper investigates the application of finite-element model updating on characterization of a nonlinear joint interface. A thin layer of virtual elements is used at a joint location to represent the nonlinear behavior of the coupling in the tangential direction. The material properties of the elements are described by a nonlinear constitutive stress-strain equation that defines the nonlinear state of the joint interface. In this study, Richard–Abbot elastic-plastic material was considered, which is capable of characterizing energy dissipation and softening phenomena in a joint at a nonlinear state. Uncertain material parameters are adjusted to minimize the residual between the numerical and experimental non...

A Tensor-Based Structural Damage Identification and Severity Assessment

Early damage detection is critical for a large set of global ageing infrastructure. Structural Health Monitoring systems provide a sensor-based quantitative and objective approach to continuously monitor these structures, as opposed to traditional engineering visual inspection. Analysing these sensed data is one of the major Structural Health Monitoring (SHM) challenges. This paper presents a novel algorithm to detect and assess damage in structures such as bridges. This method applies tensor analysis for data fusion and feature extraction, and further uses one-class support vector machine on this feature to detect anomalies, i.e., structural damage. To evaluate this approach, we collected acceleration data from a sensor-based SHM system, which we deployed on a real bridge and on a laboratory specimen. The results show that our tensor method outperforms a state-of-the-art approach using the wavelet energy spectrum of the measured data. In the specimen case, our approach succeeded in detecting 92.5% of induced damage cases, as opposed to 61.1% for the wavelet-based approach. While our method was applied to bridges, its algorithm and computation can be used on other structures or sensor-data analysis problems, which involve large series of correlated data from multiple sensors.

Adaptive One-Class Support Vector Machine for Damage Detection in Structural Health Monitoring

Machine learning algorithms have been employed extensively in the area of structural health monitoring to compare new measurements with baselines to detect any structural change. One-class support vector machine (OCSVM) with Gaussian kernel function is a promising machine learning method which can learn only from one class data and then classify any new query samples. However, generalization performance of OCSVM is profoundly influenced by its Gaussian model parameter \(\sigma \). This paper proposes a new algorithm named Appropriate Distance to the Enclosing Surface (ADES) for tuning the Gaussian model parameter. The semantic idea of this algorithm is based on inspecting the spatial locations of the edge and interior samples, and their distances to the enclosing surface of OCSVM. The algorithm selects the optimal value of \(\sigma \) which generates a hyperplane that is maximally distant from the interior samples but close to the edge samples. The sets of interior and edge samples are identified using a hard margin linear support vector machine. The algorithm was successfully validated using sensing data collected from the Sydney Harbour Bridge, in addition to five public datasets. The designed ADES algorithm is an appropriate choice to identify the optimal value of \(\sigma \) for OCSVM especially in high dimensional datasets.

Symbolic dynamics time series analysis for assessment of barely visible indentation damage in composite sandwich structures based on guided waves

This study addresses the detection and localization of barely visible indentation damage in composite sandwich structures using ultrasonic guided waves. A quasi-static loading was gradually applied on a specimen of carbon fiber reinforced epoxy with honeycomb core, with the resulting dent size varying between 0.2 and 2.7 mm. The fundamental symmetric (S0) Lamb wave mode was excited to interrogate the structure. An anomaly measure was established based on symbolic time series analysis; it was defined as the ratio between the norms of probability vectors obtained from the symbol sequence vectors before and after damage has occurred. The symbolic time series analysis method transforms time series data into symbol sequences according to a pre-constructed symbol space using a set number of partitions. The number of partitions selected was determined based on the maximum Shannon’s entropy approach. An imaging algorithm was adopted in order to localize the damage. The effects of the excitation frequency and the number of partitions on the precision of prediction were investigated. The adopted approach showed high sensitivity to a very small change of 0.2 mm on the surface panel after a quasi-static loading of 2-mm indentation. Furthermore, the ability of the method to detect progressive damage was demonstrated. The results obtained demonstrate that symbolic time series analysis has excellent potential for use in detecting small defects such as barely visible indentation damage.

Automated Operational Modal Analysis of a Cable-Stayed Bridge

11 Automated techniques for analyzing the dynamic behavior of full-scale civil structures are becoming 12 increasingly important for continuous structural health monitoring applications. This paper aims to extract the 13 structural modal parameters of a full-scale cable-stayed bridge from the collected ‘output-only’ vibration data 14 without the need of any user interactions. The work focuses on the development of an automated and robust 15 operational modal analysis (OMA) algorithm, utilizing a multi-stage clustering approach. The main 16 contribution of the work is to define a novel way of automatically defining the hierarchical clustering threshold 17 to enable the accurate identification of a complete set of modal parameters. The proposed algorithm is 18 demonstrated to work with any parametric system identification algorithm that uses the system order ‘n’ as the 19 sole parameter. In particular the results from Covariance-driven Stochastic Subspace Identification (SSI-Cov) 20 methods are presented. 21

FRF Sensitivity-Based Damage Identification Using Linkage Modeling for Limited Sensor Arrays

This paper presents a novel method to localize and quantify damage in a jack arch structure by introducing a linkage modeling technique to overcome issues caused by having limited sensors. The main...

Damage Localisation Using Symbolic Time Series Approach

The objective of this paper is to localise damage in a single or multiple state at early stages of development based on the principles of symbolic dynamics. Symbolic Time Series Analysis (STSA) of noise-contaminated responses is used for feature extraction to detect and localise a gradually evolving deterioration in the structure according to the changes in the statistical behaviour of symbol sequences. The method consists of four primary steps: (1) generating the time series data by a set of measurements over time at evenly spaced locations along the structure; (2) creating the symbol space to generate symbol sequences based on the wavelet transformed version of time series data; (3) developing the symbol probability vectors to achieve anomaly measures; (4) localising damage based on any sudden variation in anomaly measure of two adjacent locations. The method was applied to a clamped–clamped beam subjected to random excitation in presence of 5 % white noise to examine the efficiency and limitations of the method. Simulation results under various damage conditions confirmed the efficiency of the proposed approach for localisation of gradually evolving deterioration in the structure, however, for the future work the method needs to be verified by experimental data.

Structural condition assessment using entropy-based time series analysis

We present a time-series-based algorithm to identify structural damage in the structure. The method is in the context of non-model-based approaches; hence, it eliminates the need of any representative numerical model of the structure to be built. The method starts by partitioning the state space into a finite number of subsets which are mutually exclusive and exhaustive and each subset is identified by a distinct symbol. Partitioning is performed based on a maximum entropy approach which takes into account the sparsity and distribution of information in the time series. After constructing the symbol space, the time series data are uniquely transformed from the state space into the constructed symbol space to create the symbol sequences. Symbol sequences are the simplified abstractions of the complex system and describe the evolution of the system. Each symbol sequence is statistically characterized by its entropy which is obtained based on the probability of occurrence of the symbols in the sequence. As a consequence of damage occurrence, the entropy of the symbol sequences changes; this change is implemented to define a damage indicative feature. The method shows promising results using data from two experimental case studies subject to varying excitation. The first specimen is a reinforced concrete jack arch which replicates one of the major structural components of the Sydney Harbor Bridge and the second specimen is a three-story frame structure model which has been tested at Los Alamos National Laboratory. The method not only could successfully identify the presence of damage but also has potential to localize it.

On Structural Health Monitoring Using Tensor Analysis and Support Vector Machine with Artificial Negative Data

Structural health monitoring is a condition-based technology to monitor infrastructure using sensing systems. Since we usually only have data associated with the healthy state of a structure, one-class approaches are more practical. However, tuning the parameters for one-class techniques (like one-class Support Vector Machines) still remains a relatively open and difficult problem. Moreover, in structural health monitoring, data are usually multi-way, highly redundant and correlated, which a matrix-based two-way approach cannot capture all these relationships and correlations together. Tensor analysis allows us to analyse the multi-way vibration data at the same time. In our approach, we propose the use of tensor learning and support vector machines with artificial negative data generated by density estimation techniques for damage detection, localization and estimation in a one-class manner. The artificial negative data can help tuning SVM parameters and calibrating probabilistic outputs, which is not possible to do with one-class SVM. The proposed method shows promising results using data from laboratory-based structures and also with data collected from the Sydney Harbour Bridge, one of the most iconic structures in Australia. The method works better than the one-class approach and the approach without using tensor analysis.