Composite joints are widely used in industrial fields due to their advantages, such as high efficiency and low weight. However, stress concentration at the adhesive interfaces often leads to potential failure risks. To address the challenges of damage monitoring in bonded joints, this study proposes a change point detection algorithm based on the Markov Chain Monte Carlo (MCMC) method, integrated with acoustic emission (AE) techniques. Tensile tests were conducted on single-lap joints with various adhesive lengths, and the evolution of AE signal characteristic parameters under load was collected and analyzed. The study found that traditional AE indicators, such as cumulative energy, cumulative counts and the Sentry function, are susceptible to noise interference and exhibit limited capability in accurately correlating with different damage stages. In contrast, the proposed probabilistic Monte Carlo-based approach effectively captures the transitions of fracture mechanisms within the AE time series. By simulating potential stage transitions under complex probability distributions, the proposed algorithm enables quantitative identification of damage initiation and propagation stages, and successfully distinguishes the damage states of specimens with different adhesive lengths. Results from the experiments confirm that this method overcomes the limitations of traditional qualitative analyses, providing a highly sensitive and quantitative tool for load-bearing capacity assessment and health monitoring of bonded structures. This work offers significant contributions to improving the reliability of structural integrity evaluations.

Acoustic emission,Composite joint,Structural health monitoring,Markov Chain Monte Carlo,Change point detection