Abstract Summary
Railway transportation has received increasing attention recently, especially in the context of climate change due to its capability of running fully on electricity, which can be generated from green sources. With this increasing demand on railway transport, the previously considered acceptable deterioration of the infrastructure is rapidly becoming a limiting factor in its sustainable development. One such situation is encountered at transition zones in railway tracks, which are locations with a significant variation of track properties (e.g., foundation stiffness) found near man-made structures such as bridges and tunnels. These zones require maintenance more often than the rest of the track as they are prone to pronounced differential settlements. In general, the measures to counteract these differential settlements were not successful, some due to poor design or deficient implementation, but the majority due to the lack of understanding of the governing mechanisms that cause the pronounced differential settlements at transition zones in railway tracks. To identify and investigate the underlying degradation mechanisms, next to experimental investigations, researchers and engineers used computational models ranging from simple phenomenological 1-D models to complex predictive 3-D finite element models (FEM). While the latter are important in situations when specific predictions are required, the interference and interaction of multiple phenomena makes it difficult to study specific mechanisms in detail, for which the former models are sometimes preferred. Nonetheless, also these phenomenological 1-D models have drawbacks, such as the local nature of the foundation as well as the neglect of wave propagation in the underlying soil medium. This study attempts to bridge the gap between simplified 1-D models and complex 3-D FEM ones by formulating a phenomenological model in which the soil is represented by a 2-D (plane strain) continuum layer. The railway track is modelled as a beam on a nonlinear and inhomogeneous spring-dashpot layer that rests on the soil continuum and is subject to a moving constant load. The settlement and variation of track properties is restricted to the spring-dashpot layer, as the soil continuum is assumed to be linear and homogeneous. This investigation aims to shed light on the influence of incorporating the foundation nonlocality and the situations in which this is necessary for accurate predictions of the settlement. To this end, the response of the aforementioned 2-D system is analysed in depth and is compared to an equivalent 1-D system that does not incorporate the foundation nonlocality. This study can help researchers and engineers in better understanding the limitations of simplified modelling approaches depending on the specific problem investigated.
