Tool condition monitoring in milling using vibration analysis

Abstract

Monitoring the condition of cutting tools in any machining operation is very important to avoid unexpected machining trouble and improve machining accuracy. This paper presents the use of vibration analysis of the cutting process in milling to indicate the presence and progression of damage incurred by an end mill. The metal cutting experiments were performed on a mild steel workpiece without using any coolant to accelerate damage to cutter, and classical processing schemes in time and frequency domains were applied to the resulting vibrations of cutting process to obtain diagnostic information. Moreover, developing fault features were also illustrated using both scalogram and its mean frequency variation. It has been found that scalogram and its mean frequency are both capable of revealing the features of not only localized, but progressive fault more clearly in the presence of strong noise than conventional time and frequency domain analyses. Furthermore, the global average of the mean frequency variation provides a useful indicator signifying the progression of wear, whereas time domain statistics do not give any consistent trend.