本团队唐学峰副教授、2022级硕士生和川越等在Journal of Materials Processing Technology发表研究论文。

摘要：Online monitoring of defect evolution during metal forming is crucial for achieving closed-loop control of product quality. The incorporation of reinforcement phases in metal matrix composites (MMCs) results in changes to micro-defect evolution and damage modes, thereby rendering the online monitoring of defect evolution more complex and challenging. Here, the authors proposed a novel intelligent sensing approach that can not only detect the formation of micro-defect but also identify the damage mode during plastic deformation of MMCs. By leveraging anomaly detection with an autoencoder to analyze the power spectral density (PSD) of acoustic emission (AE) signals collected during plastic deformation, the signals from the TC4 matrix and TiB reinforcement in a discontinuously reinforced titanium matrix composite (DRTMC) can be distinguished. Based on the intelligent sensing framework, it was found for the first time that the evolution of the TiB signals PSD correlates with defect evolution, and TiB fractures occur during the early to mid-stages of plastic deformation. It further utilizes autoencoders in conjunction with unsupervised clustering to associate the AE signals from TiB with two distinct damage modes: fracture of TiB whiskers and microcrack penetrating the matrix. The effects of stress state on the formation of defect and damage mode were also recognized by the developed approach. The effects of TiB content and stress state on the grain-level deformation behavior and damage evolution mechanism during plastic deformation of DRTMC were analyzed by full-field crystal plasticity simulation with uncoupled damage model. A TiB content of 3% in TiB/TC4 enhances matrix slip and improves plastic deformation capability. However, under shear deformation, TiB's load-bearing contribution is minimal. High stress triaxiality from a notch causes TiB-induced cracks to penetrate the matrix at lower strains, leading to failure. This study provides a promising method for the online monitoring of defect evolution during the plastic forming and service processes of MMCs.