Surface quality monitoring based on time-frequency features of acoustic emission signals in end milling Inconel-718

Surface quality deterioration is prone to be found in Inconel-718 milling process, which leads to the decrease of fatigue life of aerospace critical safety components. Therefore, in-process detection of surface quality plays an important role in guaranteeing the workpiece quality and improving the production efficiency. In this paper, a surface quality monitoring system based on time-frequency features of acoustic emission signals is established. The main characteristic of this system is that a two-step recognition is performed through the binary clustering to realize the monitoring of surface roughness and the precise recognition of the surface defects, respectively. A new feature extraction method by means of the normalized time-frequency matrix is proposed to obtain the sensitive information of surface quality. In addition, to accomplish the clustering automatically, the improved method of clustering by fast search and find of density peaks is utilized. The experimental verification with different cutting depth is designed to test the effectiveness of the system. Validation results show that the accuracy of the two steps are 99 and 72%, respectively.