Sensoring systems and signal analysis to monitor tool wear in microdrilling operations on a sintered tungsten–copper composite material

• A sensoring system for microdrilling process of tungsten–copper alloy was studied.
• Forces and vibration signals were measured for different tool diameters.
• The correlation between measured signals and the number of elaborated holes was carried out.
• Two stages (features extraction and modeling) were used in the correlations.
• The combination of wavelet transform and adaptive-network-based fuzzy inference system (ANFIS) yielded the best results.

Monitoring of micro-scale machining processes is a key issue in efficient manufacturing. Monitoring not only reduces the need for expert operators, thereby lowering costs, but it also decreases the probability of unexpected tool breakage, which may involve damage to the workpiece or, even, to the machine-tool. Process monitoring is also of immense importance in view of the tiny tool diameters used in micro-mechanical machining. In this study, a microdrilling process was experimentally studied, which involved three different TiAlN-coated drills (diameters 0.1 mm; 0.5 mm and 1.0 mm), of a tungsten–copper alloy. Variations in tool dimensions were measured after the completion of each hole, while force and vibration signals were measured throughout the cutting process. Features were extracted from the signals by using time-domain statistics, fast Fourier transform, wavelet transform, and Hilbert–Huang transform. These features were correlated with the number of drilled holes by using statistical regressions, neural networks and neuro-fuzzy systems. The study shows that the combination of wavelet transform and neural network systems yielded the most suitable prediction of the use of tool. These results are relevant for further studies on the implementation of tool condition monitoring systems for micromechanical machining processes.