Prediction of milling force based on identified milling force coefficients and tool runout parameters in time-frequency domain

Milling force is the key variable related to machining performance. Mechanistic models are commonly used to predict milling forces. However, the identification process of milling force coefficients and tool runout parameters in this model is prone to local optimal solutions. This paper develops a time–frequency domain identification method, and then the prediction of milling force is implemented on this basis. First, a milling force model considering the influence of tool runout is established, where the milling force coefficients and tool runout parameters are identified based on the time–frequency domain identification method. The simulation and experimental data are used to compare the time–frequency domain identification method with the time-domain and frequency-domain identification methods, respectively. Then, the variation of milling force coefficients and tool runout parameters with milling conditions is studied, based on which the linear interpolation method of coefficients and parameters is used to predict the milling force. Finally, the effectiveness of the milling force prediction is verified by experiments.