Symbolic time series analysis of ultrasonic signals for fatigue damage monitoring in polycrystalline alloys

The paper presents the concept and experimental validation of an analytical tool for fatigue damage monitoring in polycrystalline alloys. Ultrasonic signals are utilized for early detection of fatigue damage during the crack initiation period. Small microstructural changes occurring inside the material during the initial stages of fatigue damage cause attenuation and distortion of transmitted waves at the receiver end. The anomaly detection algorithm is based on time series analysis of ultrasonic data and is built upon the principles of symbolic dynamics, information theory and statistical signal processing. Experiments have been conducted for both constant amplitude and block loading of 7075-T6 aluminium alloy compact specimens on a special-purpose test apparatus that is equipped with ultrasonics sensors and a travelling optical microscope for fatigue damage monitoring.