Numerical Simulation-Based Damage Identification in Concrete

High costs for the repair of concrete structures can be prevented if damage at an early stage degradation is detected and precautionary maintenance measures are applied. To this end, we use numerical wave propagation simulations to identify simulated in using convolutional neural networks. Damage subjected compression modeled mesoscale discrete element method. Ultrasonic simulation on damaged specimens performed rotated staggered finite-difference grid The ultrasonic signals used train a CNN-based classifier capable classifying three different stages (microcrack initiation, microcrack growth coalescence leading macrocracks) with overall accuracy 77%. performance improved by refining dataset via analysis averaged envelope signal. refined has 90%.