Analytical and experimental investigations of rake face temperature considering temperature-dependent thermal properties

In metal cutting, rake face temperature is a critical factor affecting the quality of processed parts and tool wear. However, this temperature is difficult to estimate because of the high strain rate and barrier effect of chips. This paper presents an integrated analytical and experimental approach to predict rake face temperature, which proposes an improved analytical model considering temperature-dependent thermal properties. The considered temperature properties mainly include thermal conductivity (TC) and thermal diffusivity (TD) of the workpiece and cutting tool, which has rarely been considered in previous studies. With the measured cutting force and tool–chip contact length, the improved analytical model can predict the tool temperature based on an adaptive method to match the thermal properties and temperature at an arbitrary point. Moreover, a relatively new in situ measurement method for directly capturing the radiation of the rake face temperature using a two-colour pyrometer and slotted workpiece is presented. This new technique can eliminate the barrier effect of the chip without interrupting the cutting process. This experimental method is proved to be able to stably measure the rake face temperature with a deviation less than 3%, furthermore, the captured temperatures are much higher when compared with conventional methods under the comparable cutting conditions. The comparison of rake face temperature between the predicted and the measured results shows an average prediction error of 8.6% of the proposed method, proving its superiority over conventional models using constant thermal properties. •An improved analytical model is proposed for temperature prediction in metal cutting.•A relatively new experimental setup is presented, which can direct access to the temperature on the rake face.•The proposed measurement method is reliable and sustainable.•The proposed analytical model works better than conventional models that use constant thermal properties.