This study systematically investigates the cracking behavior of polypropylene fiber (PPF)-reinforced recycled aggregate concrete (RAC). Specimens with coarse aggregate (CA) replacement ratios of 25% and 50% were prepared for twenty distinct test groups. Uniaxial compression tests were conducted under acoustic emission (AE) monitoring to record the peak stress and AE parameters, such as average frequency, energy, amplitude, and dominant frequency. Correlations between the mean values of AE signals and the peak stress were examined. An unsupervised machine learning technique, the k-means clustering algorithm, was employed to classify cracking modes, namely matrix cracking, fiber–matrix debonding, and fiber pull-out. The results indicate that matrix cracking exhibits low average frequency and low energy, fiber–matrix debonding is characterized by high average frequency and low energy, while fiber pull-out demonstrates high energy release. Quantitative threshold criteria for distinguishing these cracking modes are established, offering an analytical framework for identifying failure mechanisms in PPF-reinforced RAC. This analytical framework provides valuable insights for optimizing fiber-reinforced recycled concrete designs and assessing structural integrity through non-destructive monitoring.
Yang et al. (Sat,) studied this question.