The Use of Digital Image Correlation and Fiber Optic Sensing to Assess Railway Track Support Performance

Abstract

The low subgrade stiffness associated with railway corridors constructed over peat often results in large rail displacement and ground deformations during train passage, leading to accelerated track deterioration and increased risk of derailment. This thesis provides insights into track behaviour using Digital Image Correlation (DIC) and fiber optic sensing.

To measure track displacement using DIC at sites with soft subgrades a two-camera backsight method was developed to account for the movement of the camera due to ground vibrations. This method was then used to measure track displacement of the ballast crib before and after mass stabilization at a poorly performing section of track. The effectiveness of the mass stabilization was quantified using the trackbed modulus, which provides a method to isolate the improvement of the subgrade from the presence of voids between the rail, sleepers, and ballast. The results indicate that mass stabilization can be an effective rehabilitation strategy.

The use of Rayleigh backscatter fiber optic sensing was first validated in the laboratory and then used to measure rail strains and investigate the track behaviour over a 7.5 m long section of track at two different sites. Measurements from a hi-rail vehicle showed that the rail strains can be used to calculate rail curvature, which can then be used to assess the variability of track support.

The fiber optic sensing was then used at a second site where the measured rail strains, along with train axle weights, were used to infer the rail seat load–deflection relationships for individual sleepers. These relationships were observed to provide significantly more detailed information than the traditional consideration of the relationship between applied load and rail deflection.

The effectiveness of helical screw piles installed at a site with a peat subgrade was assessed based on the axial load carried by several piles instrumented with strain gauges, piezometric data, and track displacement measurements. The results of the analysis show that the piles carried a portion of the train load, however due to the seasonal variation of pore pressures and the limited track displacement data gathered it is not possible to ascertain how effective the piles were at improving the track behaviour.