Key drivers for rail track settlement and its prediction using FE model

  • Ilaria Grossoni (Speaker)
  • Jou-Yi Shi (Speaker)
  • Bezin, Y. (Speaker)

Activity: Talk or presentation typesOral presentation


Vehicle/track interaction (VTI) models can be coupled with semi-empirical formulae describing the long-term ballast behaviour in order to predict the local progressive deformation of support and the resulting track geometry change under repeated load passage. Available formulae have been mostly empirically derived for specific traffic and track type scenarios, and their applicability remains limited beyond each study they were intended for. The main aim of the present work is to pave the way for developing a generic settlement formula applicable to a wide range of vehicle and track scenarios. To do so a detailed ballasted track FE model is developed which can predict the ballast layer settlement after loading. The FE model is fully parametric so that the track support and superstructure characteristics can be varied and analysed against varying loads magnitude and vehicle configurations (axle load spacing). This way the dependency of the results on key characteristic of the track and imposed loads can be derived to drive the proposed settlement equation. This equation can then later be used in the context of vehicle-track interaction dynamic simulations on a larger scale.
The ballasted Finite Element (FE) track model was developed combining Python, Abaqus and Fortran. Porous material properties and extended Drucker-Prager model are used to account for nonlinear elasto-plastic behaviour of ballast. An elastic foundation layer has been considered under the ballast layer that emulates the subgrade response, assuming most of the settlement occurs in the ballast layer.
The initial parametric study shows that remarkable decreases in ballast settlement can be obtained either shortening the sleeper spacing either using wider sleepers, approximating the effect of a continuous support. It also confirms the widely reported benefit of increasing the resiliency of the track support using for example lower rail-pad stiffness or adding extra layers, namely under sleeper pads. The model also showed the effect of varyious types of vehicle configuration (axle spacing) and the associated variability in settlement predicted for a nominal axle pass.
Period6 Sep 2018
Event title4th International Conference on Railway Technology: Research, Development and Maintenance
Event typeConference
Conference number4
LocationSitges, SpainShow on map
Degree of RecognitionInternational