Feed rate’s effect on the stability of milling was absent in traditional two dimensional stability limit prediction diagrams. To address this issue, a new model of cutting force coefficients was proposed. This model was expressed by a quadratic polynomial equation using feed per tooth as its variable. For this purpose, central composite experiment with a set of feed rates generated by quadratic regression was designed and conducted. Milling force coefficients under certain cutting conditions were derived based on the measured milling force. The regression parameters were then estimated using least square method. The model reliability was verified by significance test. The model prediction performance was validated by comparing the prediction force with the measurement. Based on the proposed model of cutting force coefficients, the feed rate factor was introduced into the stability prediction, which converted the two-dimensional stability diagram into a three-dimensional one and revealed the nonlinear relationship among the rotational speed, feed rate and axial cut depth in the critical stability.
