Defect Detection of Ferromagnetic Rail Using EMAE-based Peak-to-peak Method and Confidence Probability Indicator

Abstract Rail defect detection holds a vital role in bolstering the safety, improving operational efficiency, and optimizing the lifespan of infrastructures in railway transportation systems. This paper proposes an electromagnetic acoustic emission-based peak-to-peak (EMAE-PTP) method along with a dedicated confidence probability indicator (CPI) for ferromagnetic rail defect detection. Firstly, a comprehensive simulation model that blends Lorentz forces with magnetostrictive effects is built up, affirming the theoretical practicability of the proposed EMAE-PTP method in ferromagnetic rail defect detection. Taking into consideration of the contingency and difference in actual signals acquisition, a special indicator, namely CPI, is formulated as the defect evaluation threshold. Based on the Chebyshev inquality and the time-domain characteristic of acquired signals, this CPI delineates the range of peak-to-peak amplitudes related to non-defective state, with a confidence level up to 96%. By doing so, the numerically segregation of defect signals can be accomplished with ease. In addition, according to the detection coefficient calculated from CPI, the suitable excitation conditions for electromagnetic acoustic emission application are determined. In conclusion, the efficacy of the proposed approach for ferromagnetic rails defect detection is substantiated, encompassing a holistic assessment of both its theoretical underpinnings and experimental manifestations.