The performance and deformation of ballast bed are significantly influenced by the particle morphology (size and shape), the rheology (translation and rotation), and the degradation (breakage and abrasion). Regarding the ballast particle morphology, the ballast particle size is generally measured by sieving and described with the Particle Size Distribution (PSD), while the particle shape is normally classified as three characteristics, the form, angularity, and surface texture. Quantifying particle morphology with current manual methods is difficult to obtain accurate results (often subjective). Concerning the ballast particle rheology, almost all the related studies are based on numerical simulations, e.g. the Discrete Element Method (DEM). A limited number of studies were performed to record the translation and rotation with the electronic devices embedded in ballast layer. However, the numerical simulations can only precisely reflect the ballast particle rheology in quasi-static tests (e.g. direct shear test), and the electronic devices can only record the ballast particle rheology in the limited areas, where they were placed. The ballast breakage could be evaluated by the change of the PSD, but the determination of PSD involves significant errors. Additionally, the manual methods could not fully quantify the ballast abrasion. As a result, more accurate evaluation methods need to be developed and utilised for the validation and confirmation of the degradation-related studies.Towards these limitations, the studies on two-dimensional (2D) and three-dimensional (3D) image analysis methods for granular materials are reviewed, discussing their existing and potential utilisation in railway ballast applications. This paper can be of interest to the researchers, who are dealing with the performance and deformation of ballast bed. Additionally, a special attention can be paid to utilising the image analysis for accurate particle morphology quantification, particle rheology investigation and ballast degradation evaluation.
- Image analysis