It has been the goal of railroad track inspectors and researchers alike to constantly strive to improve the methods being used for rail inspection. This paper proposes a method to analyze long-range rail test data with the continuous wavelet transform in order to extract reflection coefficients which may then be used to classify defects. The feasibility of using three different dynamic rail tests for defect classification of four sizes of transverse head defects was investigated. Optimum frequencies to be used for maximum sensitivity to the defects were identified.
Air-coupled ultrasonic testing is gaining increasing attention in a variety of NDE applications that can benefit from non-contact probing and ease of transducer positioning. The main focus of this paper is the assessment of an air- coupled ultrasonic method for the inspection of railroad tracks focusing on the detection of internal defects. It is shown that resonant standing waves can be successfully generated in the rail cross-section to alleviate the challenges associated with the large acoustic impedance mismatch between air and steel. The effectiveness of the resonance concept is demonstrated in through-transmission laboratory testing of the rail head, web and head flange aimed at the detection of longitudinal defects. A pitch- catch configuration potentially effective for the detection of transverse defects in the head flange is also demonstrated. The suitability of the air-coupled ultrasonic method for field use is discussed with particular attention to the achievable inspection speed. Feasibility results are also presented from the measurement of flexural (guided) vibrations in rails. These vibrations propagate long distances and could therefore be well-suited for long-range testing.