Vibration-Based Tool Condition Monitoring for CNC Grinding Process

This paper presents the development of a tool condition monitoring (TCM) model for CNC grinding machines, utilising vibration signals to extend tool life, enhance grinding performance, and ensure superior product quality. Vibration data were acquired using high-precision accelerometers during the grinding process and preprocessed to isolate relevant signal regions, reduce noise, and improve signal clarity. Key features indicative of tool wear and faults were extracted from the frequency domain using an image embedding technique. Three simplified mathematical model for tool degradation developed based on the collected dataset, with its parameters optimised using a meta-heuristic approach. Model mean squared errors for the linear, exponential, and power-law models were calculated as <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$0.0092,0.0172$</tex>, and 0.0084, respectively, with the power-law model demonstrating superior accuracy in predicting tool wear degradation. Validation with experimental data from industrial CNC grinding operations highlights the effectiveness and reliability of the proposed model in real-world applications.