Efficient simulation of the soil–structure interaction on the dynamic response of a portal frame railway bridge

The dynamic response of railway bridges can be highly influenced by the effect of soil–structure interaction. This occurs as the soil dissipates energy and modifies the flexibility of the bridge supports, which impacts the modal parameters of the structure and its response to passing trains. In the case of partially-buried structures such as portal frames, this interaction mechanism is of particular relevance. However, simulating the soil effect is complex, and may require an elevated computational effort. Under these conditions, obtaining accurate predictions of the bridge dynamic behaviour becomes challenging. For this reason, the interplay between the bridge and the soil is usually disregarded. To address this limitation, a numerical approach devoted to implement soil–structure interaction with reduced computational cost is presented in this contribution. The method is based on a substructuring scheme, and considers two numerical models: (i) a full three-dimensional finite-element interaction model, including the track, the bridge, and the surrounding soil, and (ii) a simplified version of it, in which the soil is substituted by a series of linear spring-dampers. The first model is used to derive frequency-dependent dynamic stiffness functions that describe the mechanical coupling between the bridge and the ground. Then, these functions are used to calibrate the spring-damper elements representing the soil in the subsequent simplified model, and the dynamic problem is solved by complex modal superposition. The suitability of the proposed methodology is evaluated through its application to an existing portal frame railway bridge. The effect of other relevant aspects on the bridge response such as the track irregularities and the contribution of the vehicle–bridge interaction is also taken into account. The results highlight the potential of this approach to obtain satisfactory predictions of the bridge performance in an efficient manner.