2:30 PM - 2:45 PM
[AGE34-04] Development of a numerical simulation for soil freezing processes using parameter optimization
Keywords:Soil freezing, Numerical analysis, Parameter estimation, Model validation
Water flow during soil freezing and thawing significantly influences ice formation as it can carry heat energy. This phenomenon is highly relevant to infrastructure design in cold regions, impacting road damages and the stability of building foundations. Thus, understanding the physical mechanisms of freezing and thawing is essential for maintaining infrastructure in cold climates.
Additionally, in agriculture, soil freezing affects the rhizosphere environment, significantly influencing crops overwintering and early spring growth. Its effects extend to forests and ecosystems, altering soil water supply and drainage, which in turn impacts plant growth and groundwater recharge. Therefore, elucidating and predicting soil freezing and thawing processes is essential across many different applications.
Numerical modeling of these processes is based on a coupled system of the Richards' equation and the heat convection-dispersion equation, incorporating the generalized Clausius-Clapeyron (GCC) equation to describe the phase equilibrium between liquid water and ice phases. The GCC equation accurately represents the ice-free water pressure in frozen soils as unlike empirical models, it is derived from physical principles. This makes it a highly scalable approach, allowing for analysis of different soil types using only the parameters of the water retention function.
In this study, a numerical solver for coupled water flow and heat transport under freezing conditions was developed using the GCC equation to account for phase change and was validated using experimental data. Parameter estimation for the numerical model was conducted using PEST (Parameter ESTimation) model with temperature data obtained from laboratory experiments. By optimizing parameters, the model was refined to better reflect realistic physical properties. PEST, a widely used optimization tool for numerical model calibration, improved the efficiency and accuracy of parameter estimation.
The results demonstrated that numerical simulations using the estimated parameters successfully reproduced temperature changes observed in the laboratory experiments with high accuracy. Furthermore, by incorporating PEST, the model effectively captured the detailed water flow and heat transport characteristics during freezing. These findings confirm the effectiveness of the developed numerical solver in predicting heat transport associated with soil freezing, contributing to advancements in simulation technology for soil freezing analysis and infrastructure design in cold regions.
Additionally, in agriculture, soil freezing affects the rhizosphere environment, significantly influencing crops overwintering and early spring growth. Its effects extend to forests and ecosystems, altering soil water supply and drainage, which in turn impacts plant growth and groundwater recharge. Therefore, elucidating and predicting soil freezing and thawing processes is essential across many different applications.
Numerical modeling of these processes is based on a coupled system of the Richards' equation and the heat convection-dispersion equation, incorporating the generalized Clausius-Clapeyron (GCC) equation to describe the phase equilibrium between liquid water and ice phases. The GCC equation accurately represents the ice-free water pressure in frozen soils as unlike empirical models, it is derived from physical principles. This makes it a highly scalable approach, allowing for analysis of different soil types using only the parameters of the water retention function.
In this study, a numerical solver for coupled water flow and heat transport under freezing conditions was developed using the GCC equation to account for phase change and was validated using experimental data. Parameter estimation for the numerical model was conducted using PEST (Parameter ESTimation) model with temperature data obtained from laboratory experiments. By optimizing parameters, the model was refined to better reflect realistic physical properties. PEST, a widely used optimization tool for numerical model calibration, improved the efficiency and accuracy of parameter estimation.
The results demonstrated that numerical simulations using the estimated parameters successfully reproduced temperature changes observed in the laboratory experiments with high accuracy. Furthermore, by incorporating PEST, the model effectively captured the detailed water flow and heat transport characteristics during freezing. These findings confirm the effectiveness of the developed numerical solver in predicting heat transport associated with soil freezing, contributing to advancements in simulation technology for soil freezing analysis and infrastructure design in cold regions.