Sensor fusion based technique for tool condition monitoring during milling process

In this paper a sensor fusion technique has been applied for tool condition monitoring during a face milling operation. It has been achieved by analyzing different sensor signatures obtained during the machining process and developing a neuro-estimator the condition of the cutting tool using the sensor fusion principle. A software based data acquisition system has been developed using Lab VIEW. Face milling of steel work-piece has been carried out with uncoated carbide inserts over wide domain of process parameters. Experimental trials have been carried out in the laboratory as well as in actual industrial environment. During experimental trials, different sensors were captured and the corresponding tool wear have been measured using an optical microscope. The acquired signals were pre-processed by different modules like chopping, filtering, segmentation for analyzing in different domain and extracting the features of different cutting tool conditions. These features were mapped with measured tool wear for developing a supervised neuro-estimator using different sensor feature combination. Amongst the different neuro-estimators, the force and power based sensor fusion estimated the condition of cutting tool with an error level of 34 pm. The work proposes current and voltage sensor as possible replacement for force sensor for cutting tool condition monitoring.