Intelligent identification of cracking based on wavelet transform and artificial neural network analysis of acoustic emission signals

To gain an insight into the evolution of micro-cracks in concrete materials, three typical modes of acoustic emission signals were identified by the wavelet transform (WT) of experimentally recorded AE signals. The whole damage process of a concrete structure subjected to three-point bending loading was divided into three stages: Crack initiation; crack growth; and crack coalescence, based on the event density fluctuation of the three modes of AE signals and the damage theory of brittle materials. On the basis of the distribution characteristics of the three modes of AE signals at the three damage stages and the strain release theory, AE signals were associated with crack initiation, crack growth and crack coalescence, respectively. An intelligent system of crack identification and damage evolution monitoring was established using an artificial neural network (ANN) to improve the recognition rate and the distinction accuracy; it exhibited a satisfactory performance.