Predicting the Soil Freezing Characteristic From the Particle Size Distribution Based on Micro‐Pore Space Geometry

Abstract

Abstract Knowledge of freezing characteristic of a porous media is necessary to investigate heat and water transfers in cold regions. The unfrozen water content of soil‐water system is affected by its complicated internal characteristics and external factors such as temperature and pressure. Although various experimental and analytical methods have been proposed to explore the complex relation between unfrozen water and temperature, most of them are empirical or limited. In this study, a sphere‐cylinder dual structure is proposed according to surface adsorption and double electrical layer effect, and then a theoretical upscaling framework is presented based on particle size distribution and the Clapeyron equation. The proposed model predicts the freezing characteristic in soil‐water systems from a new pore structure considering the micro‐pore space geometry. In order to evaluate the new model, 21 data sets from literature and two data sets from nuclear magnetic resonance (NMR) tests in this study are used to compare with predicted results, and excellent match is obtained. The improved method has an advantage in predicting unfrozen water content at lower temperatures. In addition, from the soil freezing characteristic curve, it can be found that the residual unfrozen water contents are positively associated with clay content. Moreover, in low temperature range, T 2 cut‐off value decreases with the decrease of temperature. As a fundamental study, the soil freezing characteristic researched in this study is helpful for the research on the hydraulic conductivity of frozen soil, which is significance for the stability analysis in cold region engineering, and the hydro‐ecological process in seasonal frozen soil area.