There is a need for novel, technology-based sensing systems and analytic methods for monitoring and improving transportation rail safety through enhanced understanding of system and component failure modes, drivers, and timing. For example, one critical need is early-stage pre-catastrophic-failure detection of rail breakage, or emerging faults under trains. While these may not lead to immediate derailment until further damage occurs at the evolving rail fracture locations, it is critical that (1) such breakages or faults be detected and (2) an alert regarding their existence be sent before catastrophic damage or derailment does occur.
The objective of this project was to develop an innovative onboard real-time broken rail and defect monitoring and detection system based on fiber-optic (FO) sensing technology that enables early-stage fault detection and repair well before the rail has reached a state which could lead to catastrophic derailment.
During the R&D work on this project at IFOS, the team exploited the broadband and high-speed capabilities of our advanced optical fiber Bragg grating (FBG) sensing and optoelectronic interrogation technology to enable real-time high-speed measurements of dynamic strain, shock, and vibration signals as well as their wideband spectral signatures as indicators of rail integrity (health) condition, both reliably and autonomously, in real time including while the train is in motion. The work comprised four technical tasks: (1) Design FBG sensor array in conjunction with data collection hardware and application-specific data analytics software. (2) Test onboard high-bandwidth FO sensor system in simulated broken rail condition scenarios. (3) Develop customized application-specific condition monitoring algorithms for broken rail detection. (4) Perform computer simulations of the train-rail system.
In this project the team defined, designed and performed lab tests to demonstrate the capability of IFOS’ high-speed broadband FO sensing system in detecting the simulated event of the train wheel hitting a breakage point in the rail. In collaboration with the project subcontractor, V3T we developed signal processing algorithms to process the onboard FBG sensor data and identify rail break events on track. We performed analytical work on studying the effect of noise, including the background acoustics and wheel-flats on the capability of the real-time FO sensing system in detecting the rail break events.
This project provided proof-of-concept of pre-catastrophic-failure rail breakage detection capability with high spatial and temporal resolutions (on the order of 1 mm at 65 mph train speed), and excellent sensing accuracy at high monitoring speeds (order of 1 Mega-samples per second or 1 MS/s, per sensor). IFOS’ proposed onboard FO sensing system provides the enabling capability to detect rail breakage where other inspection tools such as non-destructive testing (NDT), hand-held inspection, visual inspection, and signaling are unavailable, and/or impractical, and/or labor intensive, and/or of limited value. The system can augment existing track circuit-based broken rail detection. For optimal protection, the system should be installed on a data processing unit on the rear of the train. In non-signaled territory, the system can far exceed the performance of the existing methodology of visual inspection.
Next steps are to perform field testing of the on-train system.
The final report is available.
