Skill and knowledge integration for milling monitoring

This article describes a method for monitoring cutting-tools during milling processes. Its main aim is real-time building of indicators, to assess the performances of the tool/material set. It involves procedures that ensure the quality of the workpiece, optimal use of the tool and safe running of the machine. To achieve this aim, one must combine the know-hows of two specialities : mechanics and automatica. Mechanical engineers, specialized in machining, bring their extensive knowledge of the events that may disturb the machining process, while automaticians build the measuring indicators and adapt acquisition techniques to the production system. A measurement line has been implemented on a milling-machine, through which real-time acquisition supplies the significant information about the cutting process. An algorithm for calculating the specific cutting pressure and the associated variance is proposed. The data stored also allow delayed-time analysis. The specific cutting pressure is processed through a statistical method based on a mean-hypothesis test. The authors´ detection procedure provides the elements that help the operator choose the best cutting parameters. It consists in comparing the value calculated from the measures obtained at each revolution of the spindle, to a reference value. Its results can be interpreted through the heuristics supplied by a cutting expert