Acoustic emission-based classification of energy dissipation mechanisms during fracture of fiber-reinforced ultra-high-performance concrete

A neural network-based analysis method was developed to characterize acoustic emission (AE) events by the source mechanism during split cylinder fracture of fiber reinforced ultra-high-performance concrete (UHPC). Using AE tests of unreinforced UHPC and tests of individual fiber pullout, the network was trained to distinguish matrix cracking and fiber pullout. The results of the analysis showed that fiber pullout tends to dissipate more energy than matrix cracking, but there are important exceptions, and these exceptions depend on the distribution of fiber orientation inside the specimen. The AE results also showed how the energy dissipation shifts from matrix cracking to fiber pullout during damage progression.