Behaviour of Railway Bridge Transitions: New Insights From Digital Image Correlation And Distributed Fiber Optic Strain Sensing

Abstract

Railway bridge transitions are locations of abrupt track support stiffness change arising from the difference in support stiffness between the relatively softer foundation materials of the track approaching the bridge and the relatively stiffer bridge structure. This phenomenon is believed to amplify dynamic train wheel loads, which contribute to the development of persistent problematic differential track settlements that are difficult to resolve despite major research efforts by the railway geotechnics community. Novel sensing technologies such as Digital Image Correlation (DIC) and distributed fiber optic strain sensing offer a unique opportunity to further understand railway bridge transition behaviour by providing the ability to obtain simultaneous field measurements of railway track displacements and loads applied to sleepers (i.e. “rail seat forces”) for reduced requirements in terms of rail surface preparation and sensor installation time relative to other sensing technologies.

This research involved the use of distributed fiber optic strain sensors to develop and evaluate a method of calculating rail seat forces applied to the track, using measurements obtained during static train load testing in a controlled environment. The experiment allowed the presence of vertical force equilibrium between the applied train wheel loads and the resultant rail seat forces to be verified, which subsequently permitted a method of evaluating the rail seat force calculations to be developed and tested.

DIC and distributed fiber optic strain sensors were then used to instrument a railway bridge transition that was experiencing differential track settlement. The collected monitoring data, including measurements of railway track displacements and rail seat forces, were used to interpret track defects observed during a visual site inspection and determine their relevance toward the behaviour of the railway bridge transition under loading. It was found that measurements of gaps between the rail and sleeper plates could be used to obtain first-order approximations of both the shape of the differential track settlement profile and maximum rail bending stresses. The limitations of this work are discussed and potential avenues for future research in this field are presented.