Comparison of applying static and dynamic features for drill wear prediction

Abstract

This paper defines static and dynamic component parameters based on the method that converts thrust and torque detected during drilling process into equivalent thrust force and principal force. Features of the parameters are extracted by wavelet packet transform (WPT) and then used to train a back propagation neural network (BPNN) to predict the drill wear. Experiments with different drilling conditions and workpiece materials were conducted and it has been confirmed that both static and dynamic component parameters are affected by the drilling conditions. The features extracted from dynamic components in lower frequency band can predict the drill corner wear better.