Tool condition monitoring in broaching

The paper reports on research which attempts to correlate the condition of broaching tools to the output signals obtained from multiple sensors, namely, acoustic emission (AE), vibration, cutting forces and hydraulic pressure, connected to a hydraulic broaching machine. Broaching tools in a variety of conditions, namely, freshly ground teeth, one or more chipped teeth, one or more weakened teeth, one or more broken teeth and tools showing uniform wear, were investigated. Single tooth and multiple teeth broaching tools were employed during the broaching trials and the sensory signals were analysed both in the time and frequency domains. The results show that AE, vibration and cutting force signals are all sensitive to tool condition and a correlation can be made between the broaching tool condition and sensory signals using a variety of signal analysis techniques. A brief review of the advantages and the disadvantages of each sensor/signal and its associated analysis technique is presented. It is concluded that the most sensitive sensors to changes in tool condition are not necessarily those that are easy to mount or incorporate in a broaching machine. This can impose restrictions on the types of sensors that can be retrofitted to current machines. The condition monitoring system is currently being extended to incorporate sensor fusion techniques to increase the flexibility of the monitoring system and improve confidence in decision-making. (C) 2003 Elsevier Science B.V. All rights reserved.