A train accident occurs in the US every 90 minutes, many of which are attributable to undetected cracks in the tracks. Ultrasound techniques are used to test for cracks, but they are limited in application and have difficulties detecting cracks that are vertical to the top surface of the rail or that are in its base. To address these issues, researchers at Johns Hopkins University in Baltimore have developed a laser-based ultrasound system that detects flaws current instruments miss and that does so at much higher speeds.The technique employs a Q-switched Nd:YAG laser to generate ultrasonic waves in the rail by localized ablation of the surface. The laser emits 400 mJ of 1064-nm radiation with a pulse width of 4 to 7 ns, which is delivered to the rail through optical fiber or via a set of mirrors.To detect the ultrasound signal, the system incorporates a capacitive air-coupled transducer that monitors frequencies in the range of 50 kHz to 2 MHz. Although optical detection using an Nd:vanadate laser and a confocal Fabry-Perot interferometer also has been studied for the remote monitoring of high-frequency ultrasound, the transducer is sufficient for the types of cracks of concern to the railroad industry. It also does not suffer the drawbacks of optical detection; namely, its expense and the fact that its performance is dependent on the reflectivity of the surface that is under test.The hybrid system offers several advantages over current ones. With a single pulse of the laser, longitudinal, shear and surface modes of ultrasonic waves propagate in the material, enabling the detection of cracks at almost any inclination, including surface and internal cracks as well as those on the base of the rail. Furthermore, while magnetic induction systems and today's ultrasound systems require constant or near-constant contact with the sample, the new system's capacitive air-coupled transducer affords some distance between the detection system and the rail.Testing rails and wheelsThe railroad industry is funding the research, and the lab is building a prototype to demonstrate its utility in real-world tests of both rails and wheels. The latter application is a challenge, said Shant Kenderian, a researcher on the project."The wheel is turning and rolling across the track," he explained. "Sensors need to be placed so that, when the wheel is in position, mirrors and lenses come to place, and a laser is triggered. The air-coupled transducer needs to be placed so that a signal is properly detected. Nothing can obstruct the path of the wheel, and this flip-flop motion is to be repeated for every passing wheel."Kenderian is confident that he and the others can meet the challenge, and he expects to have a prototype by the end of the year.