Study on breakout detection in fast drilling EDM of film cooling holes based on acoustic signals

Ensuring full penetration during machining film cooling holes on turbine blades by means of fast ED-drilling yet avoiding back-strike are critical to the production process. This study proposes a breakout detection method for electrical discharge drilling based on acoustic signals. Acoustic signals during drilling were collected using an acoustic sensor, and variations in signal characteristics during the penetration process were analyzed. To enhance signal quality, spectral subtraction and band-pass filtering techniques were employed to eliminate noise components. The RMS time-frequency characteristics were analyzed, and two novel algorithms, the "Mean-filtered RMS algorithm" and the "RMS-based process-normalization algorithm", were introduced to smooth and normalize the RMS features for better adaptability in varying operational conditions. A multi-signal fusion cross-validation method was also developed to improve the accuracy of breakout detection. Experimental results confirmed the effectiveness and reliability of the proposed method, demonstrating its potential for real-time breakout detection in EDM drilling processes.