Temperature Distributions in Geogrid-Reinforced Soil Retaining Walls Subjected to Seasonal Freeze–Thaw Cycles

Abstract

Geosynthetic-reinforced soil (GRS) retaining walls have been widely adopted in engineering practice based on the development of several design methods. However, the temperature effect, which was proved to be an important influencing factor of retaining wall performance, was not considered in the existing design methods. This study investigates the temperature distributions of a (GRS) wall subjected to freeze–thaw (FT) cycles using (1) a lab-scale physical model test within a custom-made temperature-controlled tank; and (2) a commercial finite-element computer program for conducting numerical modeling. The numerical model was first validated using the lab-scale model test data obtained in this study, and then a full-scale numerical model was created for achieving an in-depth understanding of the temperature distribution behaviors of a real GRS wall subjected to FT cycles. Both model test and numerical results demonstrate that (1) the periodic variation of the ambient temperature during the FT process induces a temperature fluctuation in a sinusoidal shape in the soil, and the temperature distributions of the soil are distinctly related to its location inside the GRS wall; (2) the soil temperature in zones with a distance of 2.0 m to the exposed boundaries (i.e., back of facing panels and road pavement) of the GRS wall is more sensitive to the variation of the ambient temperature. Also, the temperature amplitudes in these regions during each FT cycle are greater than those in the zones far away from the exposed boundaries of the GRS wall; and (3) the frost depth in the vertical direction or frost thickness in the horizontal direction increases constantly with a decrease of the ambient temperature lower than 0°C, and vice versa. The maximum frost depth or thickness behind the GRS wall is lagged to the lowest ambient temperature and decreases with the increase of the distance to the exposed boundaries. In addition, a mathematical model was proposed for predicting the soil temperature amplitudes at different locations inside the GRS wall.