An improved quantification method for characterisation of clay microstructure using SEM

Abstract

Clay is a material widely used in environmental geotechnics. Quantification of clay microstructure properties using scanning electron microscopy (SEM) images is intuitively the simplest way, but the challenges remain in extracting more credible quantified micro-parameters that can represent the holistic macroscopic properties of clay from SEM images and with higher efficiency. In this paper, an improved quantification method is proposed to deal with this issue. The influence of different milling and drying methods on quantification results was analysed to improve reliability and efficiency. The A–K threshold determination method (A–K method) was proposed to calculate the optimal threshold in image processing. Also, the interval estimation method was introduced to obtain the optimal magnification and number of images. Then, this quantification method was applied to Hangzhou clay. The results show that anisotropy in the clay microstructure should be considered when choosing the observed surface. Moreover, there is no significant difference in micro-parameters between liquid nitrogen frozen-vacuum drying and critical point drying. A comparison among the A–K method, artificial method and theoretical methods that can determine thresholds shows that the proposed A–K method combines the advantages of all methods. The optimal magnification and number of images can be determined by mathematical theory, which also improves reliability and efficiency.