A multi-sensor fusion model based on artificial neural network to predict tool wear during hard turning

Abstract

Factors like cutting force, cutting temperature, acoustic emission signals and vibration signals can be effectively used to predict tool wear. Even though each of these factors can be used individually to predict tool wear, a more accurate prediction will be possible if these factors are considered collectively since each of these factors predicts tool wear in its own characteristic fashion. For example, high cutting temperature is an index of flank wear and crater wear, whereas variation in cutting force indicates the fracture type of tool failure more effectively. Hence a better prediction of tool wear is possible by considering the indices of tool wear collectively rather than individually. In the present work, an attempt was made to fuse cutting force, cutting temperature and displacement of tool vibration, along with cutting velocity, feed and depth of cut, to predict tool wear during turning of AISI 4340 steel having a hardness of 46 HRC using a multicoated hard metal insert with a sculptured rake face. A regression model and an artificial neural network model were developed to fuse the cutting force, cutting temperature and displacement of tool vibration signals to predict tool flank wear. The fusion model based on the artificial neural network was found to be superior to the regression model in its ability to predict tool wear.