Shang-Yu LinPo-Han ChenTay-Jyi LinPEI-ZEN CHANGWEI-CHANG LI2025-01-212025-01-21202424058971https://www.scopus.com/record/display.uri?eid=2-s2.0-85213071747&origin=resultslisthttps://scholars.lib.ntu.edu.tw/handle/123456789/724981This study presents a methodology designed to efficiently predict the tool wear curve under varying cutting parameters, eliminating the need for time-intensive experiments to collect tool wear data over the entire tool lifespan. The approach minimizes the number of required experiments by strategically leveraging the runout effect, typically considered undesirable, to introduce varying feed rates from different flutes within a single cutting experiment. Additionally, instead of considering the tool wear change rate as a function of time, the methodology converts the change rate to be a function of the tool wear, which allows ones to establish the wear change rate model based on the current tool wear but not time. In other words, unlike conventional time-domain tool life modelling methods that need to obtain the wear data at each time step across the tools´ entire lifespan, this demonstrated methodology allows for the determination of wear values across various parameters by integrating the wear change rate modeled using much fewer sampling experiments. Validation results demonstrate a root mean square error as low as 0.01 mm, underscoring the accuracy of the model. Consequently, this method offers a pragmatic means of modelling tool wear with limited data, facilitating timely predictions of the tool wear curve under diverse cutting parameters for anticipatory tool replacement and management.truerapid modellingTool weartool wear ratejournal article10.1016/j.procir.2024.10.3142-s2.0-85213071747