A robotic system for non-destructive evaluation of RCF cracks in rails using an ACFM sensor

The growing use of heavier and faster trains means that more consideration must be given to routine inspection of rails. Inspection trains, dual purpose road/track vehicles and hand held trolleys utilising ultrasonic, eddy current, alternating current field measurement (ACFM) or magnetic flux leakage sensors are the techniques which are currently employed in the rail industry. The vehicle mounted systems generally provide information on defect locations but provide less dependable data about defect sizes, which has to be ascertained by hand-held systems. However, the hand held systems require expert operators to interpret the collected data and, hence, are more costly and time consuming to use. Therefore, in order to address the requirements of higher axle weight and faster trains, the rail inspection system needs to be improved. In this research, a robotic system has been proposed for detection and characterisation of rolling contact fatigue (RCF) cracks on in-service rail tracks. The system consists of a mechanised trolley, an FS02N Kawasaki robot arm and a commercially available AMIGO ACFM system, supplied by TSC Inspection Systems. The RCF defects will be located while the ACFM sensor is fixed in position and the trolley moves at a controlled speed. The detected defects are then re-visited and the robot arm is utilised to perform a detailed scan over the defects (for characterisation) while the trolley is stationary. This paper summarises experimental results obtained when the trolley is stationary and the robot arm is used to survey the rail head. The results suggest that the proposed system has the potential to improve the quality of RCF crack characterisation and to be used autonomously to inspect the rail tracks.