Assessing tensile damage in pressure vessel steels with varying ductility using multi-feature fusion of acoustic emission signals

Acoustic emission (AE) monitoring is crucial for detecting damage and assessing the integrity of engineering structures, especially pressure vessels. However, traditional damage assessments relying solely on a single or a limited number of parameters that are highly dependent on the preset threshold, inevitably compromise the accuracy of damage identification. Therefore, this study proposes a multi-feature fusion approach, integrating threshold-independent parameters such as peak factor, entropy, and centroid frequency, to accurately evaluate tensile damage behaviour in pressure vessel steels. Results demonstrate that traditional time-domain AE parameters can effectively distinguish various damage stages of different materials during tensile testing, including the elastic deformation stage, the yield and strain hardening stage, the necking stage, and the crack propagation and final fracture stage. Notably, these time-domain parameters surge dramatically near fracture, while the centroid frequency declines significantly. Furthermore, the effectiveness of the proposed fusion parameter was compared with traditional AE parameters. Results show that the proposed fusion parameter outperforms traditional AE parameters in tensile damage stage identification, showing enhanced sensitivity in tracking micro-damage progression, particularly during the critical fracture stage. Results from this study will provide a strategy for accurately assessing the damage state using multi-feature fusion of AE signals.