ANALYSIS OF TOOL WEAR IN DRILLING USING ARTIFICIAL NEURAL NETWORK

Abstract

Tool wear in drilling is an important parameter with respect to surface quality of hole and failure of material. Operation performed with worn out tool may increase manufacturing cost. In this work, an attempt has been made to measure manually with the help of stereoscopic microscope and this result has been compared with a statiscal model in which tool wear is assumed as the function of thrust force , machining time, speed and feed. And also compared with ANN model in which input neurons are drill diameter, torque, thrust force, machining time, feed and speed etc and output is tool wear. Comparison between these three result has also been made. It is found that ANN gives best result and can be used for online tool monitoring. Experiments performed from 1 to 40th hole while drilling operations have performed on material EN-31.Monitoring of tool wear is a high robustness process and can be used in complex production system and in flexible manufacturing system.