Diagnosis of damage evolution process for asphalt mixtures using pattern recognition with acoustic emission signals. (19th April 2021)

• Record Type: Journal Article
• Authors: Qiu, Xin; Xu, Jingxian; Xu, Wenyi; Yang, Qing; Wang, Feng; Yuan, Jie
• Is Part Of: Construction & building materials. Volume 280(2021)
• Publication Date: 2021-04-19
• Subjects: Asphalt mixture -- Acoustic emission -- Fracture mechanism -- Cusp catastrophe model -- K-means; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2021.122536
• Languages: English
• ISSNs: 0950-0618

Abstract

Abstract: The fractures in asphalt materials evolving from microcracks directly affect the performance of asphalt pavements. Acoustic Emission (AE) technique plays a crucial role in dynamically monitoring the internal structural changes of stressed materials. This study attempted to characterize the fracture mechanism of asphalt mixtures by identifying the characteristics of recoded AE data. Firstly, the three-point bending tests and AE tests were simultaneously conducted to investigate the damage evolution process of asphalt mixtures through the time characteristics of AE parameters. Secondly, the cusp catastrophe model based on AE energy rate was employed to explore the mutational characteristics of AE process and to evaluate the damage state of asphalt mixtures. Finally, the k-means method was applied to distinguish the patterns of AE signals associating with the specific damage behaviors of asphalt mixtures during the flexural failure process. The results show that abrupt change of AE energy rate and cumulative AE energy could reflect the evolutionary characteristics of damage in asphalt mixtures. The mutational analysis of AE energy rate reveals that the damage state of asphalt mixtures subjected to the continuous external forces would change between instability and stability, which could be served as a complementary approach for further understanding the periodic variation characteristics of damage evolution process based on cumulative AE energy. The k-means method could effectively classify the recorded AE signals of asphalt mixtures in three-point bending tests into three distinctive clusters. The potential mechanisms for the clustered signals corresponding to different damage behaviors of asphalt mixtures could be characterized by the parametric characteristics and the wavelet power spectrum of AE signals.