MS18.4 - System Identification and Damage Detection

Structural health monitoring (SHM) is an important approach in order to evaluate the health state and so the safety and security of structures as bridges. It also helps to maintain and to extend their lives, to detect and predict their failures. This approach regroups several techniques such as vibration based inspections, acoustic, ultrasonic or magnetic field and radiographic methods. They are based on analyzing variations of some system’s characteristics and mapping them to the state of the systems and materials. For instance, in vibration based SHM, the variations of system characteristics in linear and nonlinear dynamics, e.g., frequency and damping shifts, changes of mode shapes and backbone curves, are directly related to the current state of the system. In this study, we endow an ultrasonic technique for SHM, namely Time Reversal (TR) method. This technique is based on the time reversibility of the solutions of wave equation in a lossless medium. Several experimental and numerical studies are carried out for fault diagnosis of a beam. Experimentations are based on the iterative TR process. In general terms two configurations of transducers, i.e. piezoelectric, are considered: side by side configuration and face to face. In the first one, all transducers are positioned at one side of the damage; in the second one, transducers are positioned at both sides of the damage. For both configurations, the TR is applied in three global steps: i) an excitation wave is applied via one set of array of transducers; ii) Then, signal are recorded by other set of array of transducers; iii) they are time reversed and are reinjected to the medium. If the wave emitted in step (i) is reflected by a damage, then the re-emitted wave in step (iii) converges towards the reflecting damage. The obtained experimental results are confronted by those which are detected from the finite element modeling of the system under ultrasonic excitation. Then, further investigations/protocols are proposed for fault diagnosis of such systems.

Modal parameters estimated through in field dynamic testing define the inherent characteristics of real-world structures, being therefore employed as reference information for various purposes, including the assessment of structural damage, the evaluation of operational and environmental effects, and the calibration of realistic numerical models. Among natural frequencies, damping ratios and mode shapes, the latter have been proved far more effective in localizing structural damage given their spatial dependency on the nodal coordinates of vibrating systems. Most of modal analysis applications resort to the real part of these quantities for vibration-based damage identification of structural systems, assuming them as classically damped. However, one must be aware that the classical viscous damping assumption is idealistic in case of real-world structures because they are often made up of multiple and heterogeneous interconnected parts and the damping matrix cannot be considered as proportional to mass and stiffness matrices. Indeed, the mode shapes of real physical systems are complex in most cases, and this complexity cannot be ignored as it can adversely affect the correct identification of ongoing damage mechanisms. Based on the above considerations, the present work intends to shed light on the relationship existing between structural damage and modal complexity, taking into account that the imaginary content of experimentally identified modal vectors can derive from the simultaneous occurrence of factors other than damage, such as mass loading effects, measurement noise and high modal density, among others. To this end, numerical investigations are carried out in order to track the variation of complex mode shapes in structures subjected to damage scenarios of various extent and to infer about the generalization of a damage index, recently proposed by the authors, that relies on the weighted difference of the imaginary content of complex eigenmodes to detect, locate and assess the structural damage and its evolution path.