A Feasibility Study on Damage Detection of Nonlinear Railway Ballast by Measured Vibration of In-Situ Sleeper

The ballast layer is an essential component in ballasted railway tracks. Although efficient industrial methods are available for inspection of superstructure track components, there is currently no acceptable monitoring technology for the ballast, especially under the sleepers, which is the most critical section of the ballast layer. The behaviour of ballast is understood to be very complex, with nonlinearity in its engineering properties. This paper focuses on a feasibility study on the detection of ballast damage utilizing measured vibration of in situ concrete sleeper by following the model updating method and considering the non-linear behavior of ballast stiffness. The ballast stiffness which supports track self-weight and train load is reduced in case of ballast damage, thus, enabling a means of monitoring the status of the ballast using vibration characteristics. The track’s section across a single sleeper is modelled as two spring-mass systems supported by a Timoshenko beam and an elastic foundation of spring series. For model updating, an enhanced Markov chain Monte Carlo Bayesian algorithm is adopted to explicitly address the non-uniqueness and high uncertainty of model parameters. Analysis of time-domain vibration data from the impact hammer test carried out on an indoor segment of a full-scale ballasted track with artificial ballast damage simulations showed high accuracy in detecting regions of damage and estimating the severity of damage; thus, demonstrating the feasibility of a possible extension of the proposed methodology for real-time and in-service track monitoring.