RECOGNITION OF CHATTER WITH NEURAL NETWORKS FOR TURNING PROCESS

Abstract

Chatter, ' defines as self-excited vibrations during cutting action, deteriorates Surface finish, reduces tool life, and damages machine tools. A brief study about chatter analysis which is capable of giving the mechanism of chatter in machine tool and methods for detection and prediction of chatter has been presented in the present work. A new approach chatter recognition using neural network has been developed which provides an easy implementation and speed computation. This procedure uses two synthetically trained neural networks to recognize the harmonic acceleration signals and to estimate their frequency respectively. Based on comparison of chatter frequency and natural frequency of the system, a procedure. for chatter recognition and prediction has been developed. An experimental verification has also been carried out on the ° turning process data. The importance of the recognition of chatter in machining can be seen in the quality of product and in the development of unmanned machine tool operations