Image processing method for observing ice lenses produced by the frost heave process

Abstract

During the freezing process of soil, the pore water is segregated into ice lenses with the help of continuous water migration to the freezing front. The in situ and migrated water cause significant frost heave and stress in all directions, which can threaten surrounding structures. Pore ice and segregated ice lenses directly cause changes in the frozen soil properties, such as enhanced strength and water tightness. Although some researchers have developed proper procedures to observe and quantify the formation of ice lenses in soils, the authors are going to indicate a quantitative relationship between the forming of ice lenses with the absorbed water. Therefore, a direct observation method is proposed in this study which can measure ice lenses quantitatively and in real-time, to establish a relationship between ice lenses and absorbed water. Image processing software was applied to analyze pictures of ice lenses that were taken during a frost heave test. Based on the ice lens distribution, the forming pattern of ice lens was confirmed at the macro level. A relationship between the area of ice lenses and the absorbed water was identified, which is expected to be applied for the nondestructive detection of the water content and distribution in freezing soil for real engineering.