Development of a Novel Approach to Predict Tool Life Based on Tool Path Features in Ta6V Milling

Abstract

Abstract In the manufacturing industry, titanium alloys machining has always been a challenge, mostly because of tool premature wear. Consequently, the understanding of physical phenomena and their modeling have become critical research topics for productivity improvement. The majority of tool wear model is based on cutting condition variations. However, several tool paths with the same cutting conditions lead to different tool lives. The impact of the cutting strategy is significant on tool wear and complex to anticipate. In this article, a global method allowing to estimate the impact of a cutting strategy on tool life is presented. It is based on tool path features extracted from workshop cutting power signals collection. First, optimal cutting conditions are determined according to AFNOR standard by minimizing specific cutting coefficient. Then, correlation analysis is carried out for different configurations of the database. Special iconography of correlations is used to explain links between the features. To finish, from the correlations between impact of the tool path on tool life and the features, a multiple regression model based on the method of least squares is computed to estimate the impact of the tool path on tool life. Physical correlations have been highlighted and confirmed the significant role of the cutting strategy on tool life. Models are quite accurate despite the low amount of data and the method is promising for an industrial implementation.