Advances in long train-track dynamics modelling

This paper presents recent advancements in the modeling of long train-track dynamics, specifically addressing the challenges associated with simulating long track sections. The research, initiated in 2017, has been presented at various conferences including IAVSD 2017, IAVSD 2019, and IAVSD 2021. Initially, a method for modeling three-dimensional (3D) long train dynamics was proposed and successfully achieved. Subsequently, a short track model was developed and validated, leading to further investigations for modeling comprehensive long track sections. To overcome the limitations of existing track dynamics models, this paper introduces an innovative parallel computing approach. The proposed method divides long track sections into shorter segments, which are concurrently solved using parallel computing techniques. Each shorter section is assigned to a dedicated computing core, and interaction between adjacent sections is established through boundary conditions. Additionally, a loop-track method is employed to model infinitely long track sections. The track models incorporate four layers of structures: rails, sleepers, ballasts, and sub-ballasts. The rails are modeled using the Finite Element method, sleepers as rigid bodies with six degrees-of-freedom, and ballasts and sub-ballasts as interconnected elastic blocks.