Efficiency, Feasibility, and Application of The Discrete Element Method for Ballasted Railway Track Design

Research output: ThesisDissertation (TU Delft)

56 Downloads (Pure)

Abstract

The ballast bed of a railway track consists of two granular layers, the ballast layer and the sub-ballast layer, and it is typically constructed of crushed, angular hard rocks. Studies on ballast degradation and structural performance improvement is needed to optimise reliability, availability, maintainability and safety of the railway track. Typically, the studies are conducted by laboratory/in situ tests and numerical simulations.
Within different research methods, the Discrete Element Method (DEM) is utilized to analyse behaviour of granular materials on particle level, making it suitable for railway ballast-related research. A DEM model allows for the description of interaction between particle on a mesoscopic level, while presenting the overall performance of the assembly on a macroscopic level. However, the large number of elements in a model along with the complex algorithm lead to high computational efforts, resulting in low efficiency of the DEM models. This problem limits the number of elements acceptable in a model, which means that only a limited amount of materials in a limited scale is possible to be generated and analysed. Considering the calculation time, the accepted number of elements in a mode depends on various simulated particle sizes (e.g., soil, sub-ballast, ballast) and simulated model sizes (e.g., box model, full-scale model). Additionally, it also affects the simulated loading condition, e.g. static loading or cyclic loading....
Original languageEnglish
Awarding Institution
  • Delft University of Technology
Supervisors/Advisors
  • Dollevoet, R.P.B.J., Supervisor
  • Markine, V.L., Supervisor
Award date12 Dec 2023
DOIs
Publication statusPublished - 2023

Fingerprint

Dive into the research topics of 'Efficiency, Feasibility, and Application of The Discrete Element Method for Ballasted Railway Track Design'. Together they form a unique fingerprint.

Cite this