Abstract: This paper studies the damage characteristics of fiberglass reinforced epory resin composite using acoustic emission detection technology. In this process, the sentinel function is used to characterize the degree of damage to the material, and the acoustic emission signals are analyzed through the S-transform and fuzzy C-means（FCM） clustering to obtain the damage characteristics of the material.After the three-point bending experiment, SEM（Scanning Electron Microscope） photos of the fracture surface of the test piece are taken for verification, and the following conclusions are obtained: Analysis of SEM photos reveals four damage modes: matrix cracking, fiber debonding, delamination damage, and fiber breakage; Sentinel function analysis of the entire acoustic emission event observed that the sentinel function curve showed a clear downward trend during the bending process of the test piece; The signals at different stages divided according to the sentinel function are denoised by VMD, and then the S transform is used for time-frequency analysis to obtain the characteristic frequencies of different damages, and finally verified by the FCM clustering. The results show that the sudden change of the sentinel function value can serve as an early warning signal for material fracture, and the identification of the type of material damage can be determined according to the frequency distribution results of the S transform.