• Citer

Statistical analysis of damage indicators based on ultrasonic testing with embedded piezoelectric transducers

par Deraemaeker, Arnaud ;Dumoulin, Cédric ;Dervilis, Nikolaos;Cross, Elizabeth J;Worden, Keith
Editeur scientifique Benjeddou, Ayech;Deü, Jean-François;Mechbal,
Référence (8-11 July 2019: Paris, France), SMART 2019 - IX ECCOMAS Thematic Conference on Smart Structures and Materials, International Centre for Numerical Methods in Engineering (CIMNE), page (251-262)
Publication Publié, 2019

Publication dans des actes

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

This paper aims at developing a fully automated crack monitoring system for concrete struc- tures, based on embedded piezoelectric transducers. The essential elements of the monitoring system are the embedded transducers, the interrogation unit, the post-processing method, and the determination of a threshold for automated alarm triggering. The paper starts with the description of the sensors and inter- rogation unit, followed by a presentation of the test setup which consists in a short concrete beam with a pair of embedded piezoelectric transducers subjected to a three-point bending test to induce progressive damage. Continuous monitoring is performed throughout the test and the methodology to compute a damage indicator and compensate for the acoustoelastic effect is described. The last section deals with the determination of the threshold for alarm triggering based on statistical analysis of the damage indi- cator. Two approaches are compared, the first one which relies on the assumption of normality of the damage indicator, and the second more general one based on extreme value statistics.