The geometric shape and size of sand grains are key factors that affect the mechanical and hydraulic properties of sandy soils. These morphological properties are commonly quantified via analysis of two-dimensional microscopy photos while three-dimensional measurements, such as using X-ray micro-computed tomography, are being adopted to understand and characterize the behaviours of granular materials. This study explores the relationship between two- and three-dimensional morphological properties for fragmented sand grains. Three-dimensional morphological properties of representative grains with different combinations of elongation and flatness were measured with X-ray micro-computed tomography. A framework that integrates discrete element modelling and sand grain imaging is introduced for obtaining the two-dimensional vertical projections of these grains after freefalls onto a horizontal surface. Matlab code was developed to extract the 2D silhouette outline of each grain for measuring 2D morphological properties. The relationships between 2D and 3D morphological properties were quantified, and useful empirical equations were developed. The results show that five 3D morphological properties (length, breadth, volume, surface area, and elongation) can be estimated from the 2D morphological properties of length2D, breadth2D, projection area, and aspect ratio. Estimates of 3D morphological properties including sphericity, volume, and surface area can be improved when flatness is known. A practical approach to obtain flatness from the 2D morphological analysis results is provided. The newly developed equations were evaluated using virtual soil samples and compared with previously published data for natural grains. Use of the developed equations to estimate specific surface area and permeability of soils was demonstrated.
Keywords: 2D morphological properties, 3D morphological properties, Correlation, Grain imaging, Discrete element modelling, Specific surface area