Effects of Sample Preparation Methods on Permeability and Microstructure of Remolded Loess

Abstract

To explore the influence of sample preparation methods on the permeability and microstructure of remolded loess, remolded loess collected from Heifangtai was taken as the research object. A total of 40 sets of falling-head permeability tests were conducted using two commonly used sample preparation methods, and five different dry density and four initial water content conditions. Additionally, the electrical conductivity of the leachate and the microscopic structure of the samples were analyzed. The results demonstrate that compared to the transfer wetting method, the homogeneity of samples prepared using the pre-wetting method is inferior. This difference is particularly evident when the initial water content is high. Due to the long duration of the permeability test, the pore structure is prone to change, resulting in relatively higher permeability coefficients. Moreover, the total dissolved solids (TDS) of the leachate exhibit a significant decrease with increasing seepage time, indicating the loss of soluble salts. Microscopic structural analysis reveals that samples prepared using the pre-wetting method exhibit a greater number of large pores and aggregates, which are intrinsic factors contributing to the observed differences in permeability between the two sample preparation methods. Furthermore, it should be noted that the impact of the sample preparation method on the permeability of remolded loess is more significant when the dry density is relatively low (specifically, less than 1.45 g/cm3). Conversely, when the dry density is higher, the influence becomes less pronounced.