混合型加载下钢纤维混凝土损伤过程的声发射参数分析

Using the acoustic emission (AE) technique, a series of Brazilian disk tests with a central notch (BDCN) under mixed loading are conducted to investigate the fracture mechanism of steel fiber reinforced concrete (SFRC). The evolution of the AE parameters during the fracture process is analyzed. The results indicate that the damage process consists of three stages based on the characteristics of the cumulative signal strength and load versus time relationships. The damage in the first stage is caused by the microcrack initiation, and then the coalescence and extension of microcracks. AE signals captured during the third stage are caused by the debonding and extension of steel fibers. By employing the machine learning algorithm and analyzing the AE parameters, the damage mechanism of SFRC is revealed. Using Gaussian mixture models, it is possible to classify damage sources as tensile cracks or shear cracks. Tensile cracks dominate the damage process, while shear cracks contribute to it. From the support vector machine, it is evident that the boundaries between tensile and shear cracks are not always straight lines passing through the origin.