Abstract: Based on the Markov stochastic process theory, a stochastic theoretical model for metal cutting tool wear is proposed According to this model, tool wear at any given cutting time or tool life at any given tool wear criterion should obey the lognormal distributions. Experimental results agreed well with the theoretical model, indicating that the model is practical on certain conditions. Based on the model proposed and the accelerated life test principles, tool reliability theory and the Taylor formula could be combined to make a full use of the available machining data in evaluating the reliability of metal cutting tools.

Key words: Life test, Reliability, Stochastic model, Tool wear