Non-contact ultrasonic inspection of rails and signal processing for automatic defect detection and classification

Recent train accidents, associated direct and indirect costs, as well as safety concerns, have reaffirmed the need for developing rail defect detection systems more effective than those used today. One of the recent developments in rail inspection is the use of ultrasonic guided waves and noncontact probe techniques to target transverse-type defects. A rail inspection prototype based on these concepts is under development at University of California at San Diego (UCSD). This work reports on the feature extraction and automatic pattern recognition algorithms that are being tested in the laboratory and will be added to the prototype. The results demonstrate the detection and sizing of transverse, surface breaking cracks that extend for less than 20% of the rail head cross-sectional area.