2006 | OriginalPaper | Buchkapitel
A New Damage Identification and Quantification Indicator for Piezoelectric Advanced Composites
verfasst von : Ayech Benjeddou, Sahadevan Vijayakumar, Imad H. Tawfiq
Erschienen in: III European Conference on Computational Mechanics
Verlag: Springer Netherlands
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Vibration-based damage identification and health monitoring are well established techniques for conventional structures [
1
]. However, their application for composites has attracted researches only recently [
2
]. Corresponding investigations have concerned mainly optical fibres as sensors; while the use of piezoceramics, as sensors or/and actuators, was mainly limited to Lamb waves- and impedance-based high frequency approaches. It is then the objective of the present contribution to propose an innovative low-frequency vibration-based damage identification (presence) and quantification (location, length and depth) method for piezoelectric advanced composites. It suggests, for the first time, the use of the so-called generalized or effective (structural) electromechanical coupling coefficient (EMCC) as a damage indicator instead of the classical frequency one. The EMCC is a measure of the conversion efficiency of electrical energy to mechanical one and viceversa. Since any change in the host structure stiffness, due to the damage, considerably affects the energy conversion of the piezoelectric devices, the structural EMCC is shown to be a good candidate as a damage indicator. A structural EMCC change factor (ECF), from the healthy to the damaged state, is introduced for the damage identification and compared to the corresponding classical frequency change factor (FCF). It is found that the magnitudes provided by the former are higher than those from the latter. Hence, they can be used efficiently in artificial neural network (ANN) nonmodel based damage quantification in piezoelectric laminated beam structures [3]. Parametric analyses on the damage variables under various mechanical boundary conditions are conducted for the education of the ANN. It is found that the damage can be quantified within a maximum error of 6.5%. It is also evidenced that the ECF predictions are much better than the FCF ones.