Determining optimal replacement time for metal cutting tools

Traditional tool life models do not take into account the variation inherent in metal cutting processes. As a consequence, the real tool life rarely matches the predicted values. To compensate for this uncertainty, tools are usually replaced prematurely, which leads to unnecessarily high tool costs. In some cases, however, wear-out occurs earlier than predicted, which imposes a risk of workpiece damage or rework and can lead to other extra charges. To balance these costs, this paper proposes an age replacement model. It is assumed that penalty costs are incurred each time a tool fails before the planned replacement. The probability of such an event is determined from the tool reliability function, which models the wear-out by a mixture of Weibull distributions, while failures due to external stresses are accounted for by a homogeneous Poisson process. The optimal replacement time is then determined from a total time on test (TTT) plot. The adequacy of the proposed approach for practical application is tested and confirmed in a case study on turning of Inconel 718 with cubic boron nitride (CBN) tools.