At first glance the operational performance of ballast appears trivial in its simplicity. However, various mechanisms affect the short and long term performance of the ballast and in turn, the response of the track structure, both on a macroscopic scale considering the track structure as an entity and on a discrete particle level or mesoscale. The nature and geometry of the material itself creates difficulty in instrumenting the ballast directly and analytical solutions of the track structure in three dimensions require complex numerical models. The importance of experimental studies to establish the influence of the granular fabric on both the static and quasi-static deformational response of granular materials has been repeatedly highlighted by other researchers. A three-dimensional, half-space model is extremely difficult, if not impossible, to be developed for the theoretical investigation of complex ballast vibrations. This information contributes to the fundamental understanding of mechanical behaviour of granular media.
The proliferation of low cost Micro-Electro-Mechanical Systems (MEMS) has evolved the science of metrology through the introduction of a new class of instruments, known as “smart ballast”, for use in railway and transportation research applications. These instruments are able to quantify both the micro- and macroscopic dynamic response of railway ballast. Such a smart ballast prototype, named Kli-Pi, has been developed by the author. The name is derived from the Afrikaans word “klippie”, which is synonymous with the description of “small rock”. The instrument provides sufficient resolution and fidelity to investigate the relationship between ballast translation and rotation in three spatial dimensions for both laboratory and in-service track environments. Using multiple Kli-Pis
arranged in succession, the dynamic performance and characteristics (for both the time and frequency domain) of ballast in both laboratory and field environments can be remotely monitored in real-time. Statistical parameters and energy metrics are formulated to compare and quantify the measured laboratory and field characteristics of ballast.
Through this research, Kli-Pi has been developed to provide valuable information surrounding the rotations and deflections in three dimensions, dominant frequencies and harmonics of the track components, indirect observations of principal stress rotations (PSR) and the underlying probabilistic micromechanics and statistical nature of the mesoscale ballast dynamics. Contrary to traditional deterministic mechanics, the probabilistic nature of the material does not conform to uniform-strain assumptions. Probabilistic behaviour governs the particle deflections where instances of alternating positive and negative strain are observed in the vertical direction. For both the laboratory and field experiments investigated, the largest component of the energy was concentrated along the vertical direction, parallel to the direction of the load. The passage of the locomotives coincided with significant longitudinal and lateral forces from the tractive effort, the influence of which extended to the subballast interface. Kli-Pi provided the necessary sensitivity to observe changes in the skeletal structure or “fabric” of the ballast and the influence of impact loads. Finally, the kinetic energy of each Kli-Pi was quantified which relates statistical parameters such as standard deviation and skewness with the particle coordination number (CN) and relative degrees of confinement. Of all the statistics investigated to quantify the ballast dynamics, the mechanical work proved to be the most suitable descriptor, exemplifying the fundamental, probabilistic response of ballast subjected to dynamic loading conditions.
Dissertation (MEng)--University of Pretoria, 2019.