Tool condition monitoring (TCM) is critical for micro-machining brittle materials to ensure precision, extend tool life, and maintain surface quality. This study investigated the integration of acoustic emission (AE) sensors and cutting force dynamometry for real-time monitoring of tool wear during micro-milling of glass and silicon substrates. A total of 150 slots were machined using 0.9 mm diamond-coated micro-end mills across progressive stages of tool wear, with separate tools used for glass and silicon substrates. Data were collected across three tool wear states: Initial Wear, Medium Wear, and Severe Wear, with approximately 25 slots machined at each wear stage per material. Tool wear progression was quantified using cutting-edge radius (CER) measurements obtained via SEM imaging, ranging from 7 μm (new tool) to 23 μm (severe wear) for glass and 7 μm to 21 μm for silicon. During each machining operation, signals were simultaneously acquired from acoustic emission (AE) sensors (100 kHz sampling rate) and a cutting force dynamometer (40 kHz sampling rate). Multi-domain signal processing, including time-, frequency-, and wavelet-domain analyses, was applied to extract diagnostic features. A Random Forest classifier trained with Leave-One-Out Cross-Validation (LOOCV) achieved 92% classification accuracy, demonstrating robust generalization across tools. Results revealed strong correlations between increased cutting forces, elevated AE spectral energy (4–6× baseline in high-frequency bands), and tool degradation. AE sensors demonstrated superior sensitivity for early-stage wear detection, while force signals provided reliable indicators during medium and severe wear stages. The integrated multi-sensor framework offers substantial improvements over single modality approaches and provides a foundation for real-time TCM implementation in precision micro-manufacturing environments.
Abu et al. (2026) studied this question.