A Hidden Markov Model Approach to the Study of Random Tool Motion during Machining

Abstract

This paper presents a new approach to the study of random tool motion during machining. Theory of the hidden Markov model is applied to formulate a comprehensive random excitation system present during machining. Based on the microstructural analysis, characteristics of the hardness distribution in the material being machined are identified for analyzing the cutting dynamics in microscale. The machining action within one revolution of the workpiece and the relation between the machining actions in consecutive revolutions are interpreted as a double stochastic process. Computer simulation based on the hidden Markov model approach is used to predict values of surface roughness characterization indices under given machining conditions. The predictions are compared with the data obtained from direct measurements, showing good agreements. The developed approach has brought new light on a better understanding of vibration control during machining.