Real-Time Monitoring of Fatigue-Induced Damage in Stainless Steel through Acoustical Emission and ML Techniques

Abstract Engineering machine safety is greatly dependent on fatigue crack structural health monitoring (SHM). The acoustic emission (AE) method is a tool used for structural health monitoring (SHM) that can effectively track the real-time fatigue cracks growth (FCG). The objective of this work was to analyze the evolution of fatigue-induced damage in stainless steel (SS) utilizing AE and machine learning methodologies. The acoustic emission data acquired at the entire fatigue-loading process were gathered and examined in correlation with the advancement of damage caused by fatigue. The AE-source processes were examined through the examination of waveform features and the use of bispectrum analysis. Furthermore, a range of machine learning algorithms were utilized to identify various levels of exhaustion, and the results indicate the effectiveness of using machine learning algorithms to categorize different substages of weariness. This research is distinctive since it employs machine learning techniques to categorize different phases of weariness. This approach diverges from the current paradigm, which depends on former data of acoustic emission - loading past data and the evaluation of △K.