Computational Railway Dynamics - Integrated Train - Track - Subgrade Modeling and Simulations
Doctoral thesis, 2007
The primary objective of the thesis has been to extend the capabilities for computational modeling, prediction and simulation of the elastic response of the train, track structure and soil in railway systems. The industrial background is the high level of ground borne vibrations reported from citizens as a consequence of train traffic (in particular from high speed trains) in several Swedish locations. Computational analysis is a tool for enhanced understanding of the governing phenomena, preliminary tests of different countermeasures and for visualization of the behaviour of the system.
An advanced application encompassing finite element analysis, a large variety of solvers for large systems of equations and generation of output for graphical visualisaton was developed prior to this project. During the course of the work described in this thesis, it has been enhanced by three main features:
The integration of three-dimensional rigid body dynamics with finite elements for analysis of vehicle / solid interaction problems
The insertion of a fictitious layer of viscoelastic elements of a viscoelastic material to reduce unphysical reflections in the solution
Moving mesh calculations allowing for analysis over long distances
The outcome is an advanced application for vehicle – solid interaction problems over long distances. The inclusion of rigid body equations incorporates the dynamic behaviour of the train components, allows for tests of various passive or semi active suspension strategies and makes it possible to analyse the results in the carbody, bogies and wheels. Wheel – rail interface modelling accounts for contact stiffness theory and opens up for simulations of the effect of various types of imperfections. Usage of a moving mesh extends the capacity to selectively relocate a predefined mesh so that the track / ground geometries and track design can be modified as a function of the traveled distance.
The high level of detail leads to a system with many degrees of freedom. It is of great importance to consider the relationship between the range of problems that can be solved, the computational resources and efficiency of the numerical procedures and the chosen model. This is one of the main themes of the appended articles.
Simulations of real world problems confirm that passage at certain critical train speeds leads to strongly magnified vibrations in the track and soil. In another study, mitigation techniques show that lime cement columns can provide very good protection against ground borne vibrations. Finally, it has been shown that rail head irregularities beyond allowed levels can induce vibrations threatening the comfort of the people in the coach.