UC San Diego

Between January 2002 and January 2012, over 6,500 derailments due to internal defects and track misalignments, known as sunkinks, occurred in the US, with associated ̃ 1 billion dollars of direct and indirect cost to the railroad industry. Internal defects in rails are the result of manufacturing defects and rail wear. Current rail inspection methods have well known limitations that prevent them from detecting all of the critical flaws with minimum false positive rates. Safety Recommendations issued by the National Transportation Safety Board (NTSB) following the disastrous train derailments at Superior, WI in 1992 and Oneida, NY in 2007, among others, reiterated the need to improve current rail inspection methods. Sunkinks are the result of excessive compressive thermal loads developed in Continuous Welded Rail (CWR) in hot weather. In order to monitor the possibility of a sunkink, the rail industry necessitates of a reliable method to indicate the level of thermal stress in the rail in a nondestructive and practical manner. Both the Federal Railroad Administration and the railroads are still searching for such method for in-situ thermal stress measurement in rails. is not the search. The goal of this dissertation is to improve the state of the art of ultrasonic wave propagation in waveguides, with specific applications to : (1) improving internal defect detection in rails, and (2) measuring thermal loads in rails. On the first application (defect detection), a prototype has been developed using non-contact means of generating and detecting guided waves in rails. Specifically, the prototype uses a pulsed laser for wave generation and air- coupled sensors for wave detection. The system also employs real-time statistical processing of the ultrasonic measurements that maximizes the sensitivity to defects while minimizing false positives. The prototype was field tested with excellent results in terms of defect detection reliability. On the second application (thermal stress measurement), the techniques of guided-wave velocity change and Electro-Mechanical Impedance (EMI) changes have been explored. The wave velocity change is traditionally used in the field of bulk waves (acousto-elasticity). The extension to guided waves is not obvious. The EMI method is conventionally used to detect structural defects close to the probing transducer by measuring the coupled transducer-structure electrical impedance. The EMI method is less used to measure load levels in the structure. A model has been developed to predict the electro-mechanical response of a piezoelectric transducer mounted on a structure subjected to quasi-static loads. Experimental validation has been performed at UCSD's Large-Scale Rail Buckling Testbed, a unique 70ft-long track of CWR that was constructed at UCSD's Powell Labs for this project. Both techniques of guided-wave velocity measurements and EMI measurements have shown to have some promise, once the effects of temperature alone are compensated for