Experimental Study of Static and Dynamic Loading Effects on Guided Wave-Based Intelligent Monitoring of Welded Rail Joints

A serious concern in high-speed railways' operation and maintenance is crack initiation and propagation in welded rail joints. Piezoelectric sensors, which emit and receive ultrasonic guided waves (UGWs), can be useful in a real time intelligent monitoring of these critical spots. However, such smart sensors are highly sensitive to environmental and operational conditions. Herein, the impact of static and cyclic loads on guided waves propagating in a welded rail specimen was experimentally investigated, and the difference between damage and loading effects was discussed. Twelve PZTs were mounted on different parts of the rail section, including the rail web and top and bottom of the rail foot to generate or receive UGW signals. Time-frequency analysis and time-domain signal processing were used to demonstrate the signal changes under loading conditions. The group velocity of various parts of the rail section remained almost constant for frequencies above 300 kHz. The time of flight (TOF) of the second wave packet measured by the straight transducer pair was a reliable indicator of damage to the rail edge. According to signal processing results, the short-time temporal coherence feature extracted from the oblique transducer pair was effective in discriminating structural damage from various loading conditions.