REMAINING USEFUL LIFE ESTIMATION FOR CUTTING TOOLS BY STATISTICAL TREND EXPLORATION AND STOCHASTIC MARKOV APPROACHES

Abstract

This thesis deals with development of methodology to obtain Remaining Useful Life (RUL) of a turning tool using two different approaches; statistical trend exploration technique and stochastic Markov method. An experiment is performed by HSS tool to machine a mild steel work piece for a constant length on lathe machine. Flank wear width for different number of passes of tool over work piece is recorded for constant feed, speed and depth of cut up to failure value of tool flank wear i.e. 0.5 mm. In statistical trend exploration technique, the behavior of flank wear width is plotted on different curve against number of passes. Best fitted curve behavior is studied and equation of that curve is generated which gives us RUL for different number of passes of tool as input value. In stochastic Markov method, a state based model is developed considering four gradually degraded stages of the tool. The degradation rate among states are obtained from the data of the experiment. The rate equations are derived for the four state Markov model representing the change of the state probability with respect to time for each state. The set of equations is solved analytically by Range-Kutta method using MATLAB software. The analytical results are verified by Monte Carlo Simulation. The estimation of remaining tool life is important in planning condition based maintenance program and helps us by preventing any loss during production.