An advanced approach for coupled train-track-soil vibration analysis using impulse response functions and finite element method

Abstract

The commonly used two-step modeling approach for studying railway-induced ground vibrations involves first establishing a source model and then using the finite element method (FEM) for secondary modeling. In this process, the wheel/track interaction forces or tie/soil interaction forces from the initial step are applied as excitations to compute wave propagation in the surrounding soil. However, a key issue with this method is that the two mechanical models created for the same geotechnical site may not be equivalent. While they may be equivalent under static conditions, this equivalence may not hold under dynamic conditions. Specifically, the dynamic response of the soil to excitation in the second step could differ from the predictions made by the source model, potentially leading to discrepancies in the simulation results. This raises concerns about the accuracy of the two-step modeling approach in capturing the true dynamic behavior of the soil and its interaction with the railway system. To overcome the limitations of the two-step modeling approach, this study utilizes the impulse response function of the ballast and underlying soil to derive the dynamic compliance of the ballast. This dynamic compliance effectively represents the interaction between the ties and the soil, enabling the development of a time-domain mathematical model for the coupled vibration of the train, track, and three-dimensional finite element soil system. Additionally, an iterative calculation scheme for the numerical solution of the system of equations is introduced, facilitating a stepwise approximation approach. From a mechanical standpoint, this approach is more rigorous and logical than traditional time-domain methods. By reducing the computational domain length of the finite element model in the track direction, this method improves computational efficiency while maintaining high accuracy. Overall, this novel modeling technique enhances the accuracy and efficiency of simulating the dynamic behavior of railway systems, especially in geotechnical contexts.