Tool wear analysis of titanium alloy parts based on multi-mode perception fusion

Tool wear directly affects the processing quality and the service life of milling cutter. However, the existing methods are difficult to achieve effective measurement of the wear of milling cutter, which brings great quality risks to aircraft manufacturing. In order to realize the direct and effective online measurement of tool wear, a algorithm is proposed, which mainly includes the recognition and acquisition modal feature images based on convolutional neural networks, the tool wear analysis strategy based on multi-modal perceptual fusion to identify the damaged area, the specific quantitative value of tool wear based on multi-modal fusion analysis, and the tool wear evaluation based on the fusion results. Four titanium alloy milling cutters were tested. The proposed method can accurately identify defect areas, and the max standard deviation only accounts for 1.66%, 3.52%, 2.57% and 2.04% of the standard value. The results show that the proposed method has good wear loss perception fusion analysis capability, and the combination device can realize online accurate measurement of wear loss, laying a foundation for engineering application.