MONITORING OF TURNING OPERATION VIA FORCE SIGNALS .1. RECOGNITION OF DIFFERENT TOOL FAILURE FORMS BY SPECTRAL-ANALYSIS

During machining, the cutting edge is subjected to various forms of tool failure such as: progressive wear leading to plastic deformation; chipping; fracture or breakage. Force signals are highly valid carriers of information about the machining process and they are, hence, among the best alternatives for tool monitoring and diagnostic techniques if a proper manipulation technique is devised. A cumulative power approach using the cumulative distribution function (CDF), is introduced in this study to investigate the possible correlation between the force signals variations and the different forms of tool failure in turning operations. The CDF, in association with the proposed failure index, has added some advantages over the simple computation of average forces. It is more accurate and sensitive (better signal-to-noise ratio) approach for detecting small disturbances resulting from any of the various modes of tool failure. Also, the technique has the ability to distinguish between various forms of tool failure and this allows for better judgement of the actual tool state. Additionally, the approach is proven to be less sensitive to cutting conditions and therefore produces a more universal solution to the problem of tool wear monitoring and control.