Estimating the Reliability of the Inspection System Employed for Detecting Defects in Rail Track Using Ultrasonic Guided Waves

Abstract

AbstractThis work focuses on the implementation of a data-based method to determine the inspection system reliability in terms of detecting different types of damages in rail tracks using ultrasonic-guided Rayleigh waves and a probability of detection (POD) technique. In this study, the reliability is tested against a surface crack (SC) and sub-surface damage – a through-side thickness hole (TSTH). The guided Rayleigh waves are generated using a custom-designed sensor that excites Rayleigh surface waves in the specimen and the propagating waves are sensed on the rail track surface. The wedge shape design of the sensor helps to excite a specific ultrasonic mode in the sample thereby hindering the ultrasonic energy of other coupled guided waves that can propagate simultaneously and the wedge angle is determined according to Snell’s law relying on the wave velocity of Rayleigh wave and bulk longitudinal wave. The guided wave responses as a function of varying severity of defects are obtained through a simulation study after the verification of the obtained guided wave responses with the help of an experimental study. A damage index (DI) is defined depending on defect size that gives the trend of damage severity from the captured ultrasonic responses and for monitoring defects in the rail track. This DI is eventually fed into the POD model to determine the probability of defect detection which in turn helps determine the inspection system reliability. The POD method also helps to study the critical design parameters that could affect or improve crack detection results.Purpose – To determine the reliability of inspection system deployed for interrogating health status of rail track.Methodology – Employing the Probability of detection technique for determining how reliable the inspection system is in detecting the health status of the rail track specimen using the ultrasonic guided waves.Findings – It has been found that the proposed inspection system is >90% reliable in detecting defects.Implications – This methodology can help maintenance engineers to make an informed decision on their developed technique for investigating the health status of the rail track sample.Originality/ value – 13%.