The long-term goal of this research is to develop a continuous, non-contact rail monitoring system for high-speed defect detection. Railroads play a vital role in the safety, prosperity, and well-being of our communities and businesses. However, due to increases in train speeds and axle loads that are being applied to aging and deteriorating railroad tracks, the occurrence of broken rails has become of increasing concern for owners, and regulators. Current inspection methods, including ultrasonic testing (UT), magnetic flux leakage, and eddy current testing, have limitations in speed, coverage, and detection accuracy. UT, for instance, requires fluid-coupled transducers and examines discrete rail cross-sections, limiting efficiency and real-time defect detection. The proposed technology consists of a non-contact laser Doppler vibrometers (LDVs) system mounted on a rail car. The system will continuously capture rail vibrations generated by wheel-rail interactions as the train travels at speeds between 30 and 80 mph. We hypothesize that these vibrations can reveal internal rail flaws through detectable changes in the structural response.
In this project we plan to conduct preliminary studies to investigate the feasibility of the proposed approach. Specifically, small scale laboratory tests will be carried out on rail samples with known defects. These specimens will be mounted on ties and ballast to replicate real-world conditions, and impact hammer tests will simulate wheel-rail contact excitations. Advanced signal processing techniques will be developed to improve accuracy and minimize false positives of the proposed approach.