3D Analysis of gravel surface texture

Abstract

Surface texture plays an important role in understanding the mechanical and hydraulic behavior of a soil but has drawn little attention, especially for three dimensional parameters. Currently, the few studies on measurements and quantification of surface texture are limited to sand sized particles. To date, the surface texture of sand has been successfully obtained by using a high-resolution optical microscope for a given measurement area that is small and comparable to the sand particle size. However, for larger particles, such as gravel, the larger surface area may create difficulties. For instance, as the particle size increases, the surface texture could be influenced by other factors such as the mineral composition, scale-dependent fabric and others, hindering the simple extension of a method that quantifies the surface texture of sand to gravel. This paper attempts to address this challenge by proposing a method to measure and quantify the surface texture of gravel which explicitly takes into account the different texture scales. Crushed granite is taken as the testing material. To explore the surface texture at different scales, a 3D laser scanner (at millimeter scale) and a high resolution optical microscope equipped with interferometry (at micro-meter scale) have been employed. The novel method defines and quantifies the surface texture of gravel with the aid of the power spectral density function (PSD) and a fractal method. It was found that the fractal dimension at the two different scales do not coincide, which might be interpreted as two different self-affine patterns of the surface. For example, the features of single minerals (e.g. fractures, cleavage, hardness) dominate at the small scale and become less prominent at the large scale where grain size and shape prevail. When a single fractal dimension from PSD at the large scale is assumed for both scales, the associated error can be quantified using the square root of the surface height to a mean plane and is found to be within 15%. This suggests that the surface texture examined at a large scale by using the 3D laser scanner could be representative of the gravel, though cautions should be taken in individual cases for any quantitative studies of the effect of the surface texture.