WAVELET ENTROPY-BASED DAMAGE IDENTIFICATION TECHNIQUE FOR HYBRID FRP-CONCRETE STRUCTURES

摘要

Civil infrastructure, particularly bridges, is designed and built to be safe against failure and to perform satisfactorily during their service life. However, this infrastructure has been deteriorating at an alarming rate due to aging, inadequate maintenance, adverse environmental conditions, and constantly growing transportation demand. Utilization of corrosion-resistant composite materials, particularly fibre-reinforced polymers (FRP), has been an encouraging solution for the durability problems of concrete bridges. However, similar to conventional structures, the performance of hybrid FRP-concrete structures can be affected by various types of damage. Therefore, in order to ensure both safety and serviceability of these structures, it is essential to identify damage at the earliest possible time. In this paper, a damage identification technique (DIT) capable of detecting and localizing structural damage, and estimating its severity is presented. The DIT is based on the following: (1) structural damage changes the energy distribution of the acceleration signals of structural components under ambient vibrations; (2) these changes are detectable by means of discrete wavelet transforms (DWTs); and (3) the detected changes can be quantified using spectral entropy. The efficiency of the DIT is illustrated experimentally on a hybrid FRP -concrete bridge truss girder tested under static loading up to failure. The truss girder consists of pretensioned top and bottom concrete chords connected by precast web elements made of glass-FRP (GFRP) tubes filled with concrete. The results have demonstrated that the wavelet entropy-based DIT is able to detect damage in hybrid structural elements and is capable of localizing the damage and estimating its severity.