The delay-and-sum (DaS) methodology is implemented in the current study to depict the probability of damage presence on an aluminum plate. Compared to previous researches, the introduced modification in this work attempts to decrease the dependency of the method on the chosen frequency. Firstly, the propagation of the Lamb waves in the plate is modeled numerically employing commercial finite element package, ABAQUS. Additionally, an experimental setup is built, in which piezo patches are utilized to excite the Lamb wave and to measure the corresponding response on the plate as well. Analysis of the experimental results shows the feasibility of the method for damage detection.
Identification of damage in its early stage can have a great contribution in decreasing the maintenance costs and prolonging the life of valuable structures. Although conventional damage detection techniques have a mature background, their widespread application in industrial practice is still missing. In recent years the application of Machine Learning (ML) algorithms have been more and more exploited in structural health monitoring systems (SHM). Because of the superior capabilities of ML approaches in recognizing and classifying available patterns in a dataset, they have demonstrated a significant improvement in traditional damage identification algorithms. This review study focuses on the use of machine learning (ML) approaches in Ultrasonic Guided Wave (UGW)-based SHM, in which a structure is continually monitored using permanent sensors. Accordingly, multiple steps required for performing damage detection through UGWs are stated. Moreover, it is outlined that the employment of ML techniques for UGW-based damage detection can be subtended into two main phases: (1) extracting features from the data set, and reducing the dimension of the data space, (2) processing the patterns for revealing patterns, and classification of instances. With this regard, the most frequent techniques for the realization of those two phases are elaborated. This study shows the great potential of ML algorithms to assist and enhance UGW-based damage detection algorithms.
Vulnerability of structures to damage during their service time brings up the necessity of design and implementation of an intelligent procedure to assure the health of the structure. In the sight of this requisite, current work deals with extending the capability of a dual Kalman filter (DKF) state estimation scheme to assist vibration-based health monitoring methods. This is met by estimating the response of the structure for locations at which a sensor cannot be placed. The capability of the DKF method in the estimation of states of a linear system with an unknown input has been presented in various recent works. In this paper, a DKF approach incorporated with a reduced order structural model (in this case an aluminum plate) is utilized to obtain an estimation of applied force and the response of the structure in terms of acceleration, velocity, and displacement. These estimations are based on measured accelerations at a limited number of points on the aluminum plate as well as the state-space model of the dynamic system. Numerical simulations and experimental works are performed to obtain the mentioned datasets. To assess the robustness of the method concerning various conditions, the effect of the frequency, as well as type of the function of the input force on the validity of the method, is presented. Moreover, it is shown to what extent the number of selected modes in model reduction procedure can influence the accuracy of the DKF technique.
Ultrasonic guided waves are recognized as one of the promising approaches to detect damage in the structures. Furthermore, machine learning algorithms with their distinguished capability in recognizing and classifying the patterns in a dataset can be employed as an extension to conventional damage detection techniques. In the current study, a damage detection framework is proposed, which is based on the result of the numerical simulation of wave propagation in an aluminum plate. The feature vector required for the training of the classifiers is generated by employing statistical analysis of the time domain signals as well as their Hilbert transform. This study shows the importance of the selection of the right combination of the features for the classification of the state of the structure. The result of this study contributes to devising a more automated and generalized process for damages detection in metallic structures.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.