11:00 AM - 11:15 AM
[AGE28-07] Evaluation of saturated ground freezing analysis methods with temperature recovery method
Keywords:Ground Freezing, Numerical analysis, Temperature recovery method
To ensure the effectiveness and cost-efficiency of the ground freezing technique while addressing environmental concerns, it is essential to have a model that can predict changes in ground temperature during freezing well. There are some difficulties in simulating the ground temperature during the ground freezing processes because of the release of latent heat release during freezing. The formation of ice in soil restricts water movement, promotes increased soil stability, and enhances the durability of geotechnical structures. However, the continuous influx of heat from upstream groundwater flow poses an obstacle to the formation of frozen ground. Numerical analysis is required when advection terms are present to obtain the temperature distribution over the entire ground, as an analytical solution is not available. This study aims to develop a numerical solver that uses temperature recovery method (TRM) to account for the release of latent heat. Additionally, the study evaluates a method of directly accounting for latent heat in the numerical analysis of the freezing process. The validity and performance of the developed solver were verified by comparing the simulated results with analytical solutions.
The governing equations of the heat transport in the ground considered conductive and convective movement of heat energy. For TRM, the latent heat release is not considered in the governing equation.TRM is a numerical approach used to treat latent heat associated with freezing and thawing of water. In this method, temperature analysis is performed at each time step without considering the release of latent heat. If the temperature goes below the freezing point at a given time step, the temperature is brought back to the freezing point by assuming that the amount of temperature decrease from the freezing point is equal to the amount of latent heat released. The simulation continues until no water remains unfrozen. In another approach, the release of latent heat during the phase change between the liquid and solid phases is directly considered in the governing equation accounting for conduction and convection.In this approach, either the unfrozen water content (thus ice content) is obtained by using the General Clausius-Clapeyron (GCC) equation or the ice content is modeled as an exponential function of temperature in so-called the power model. The finite element method was used to discretize the governing equation to simulate temperature changes in a two-dimensional domain with a size of 2 m by 0.1 m. One end of the domain was kept -20 ℃ while the initial temperature of the domain was 18 ℃.
The comparison between the numerical solution using TRM and the analytical showed that the TRM method achieved high accuracy. The TRM result however showed stair-like fine stagnation during the temperature drop. As the GCC model and the power model revealed a sudden and discontinuous increase in the apparent heat capacity with the onset of freezing due to latent heat, the numerical solutions can be unstable and challenging. Compared with the TRM, both approaches resulted in much slower temperature changes. The impact of latent heat considerably decreases the apparent thermal diffusivity α below 0 ℃, slowing down the temperature decrease and leading to a large difference from the results of the TRM.
TRM was effective for ground freezing analysis and achieved higher accuracy than the methods incorporating latent heat directly into the governing equations. Differences in the modeling of thermophysical properties were found to have a significant impact on the freezing analysis.
The governing equations of the heat transport in the ground considered conductive and convective movement of heat energy. For TRM, the latent heat release is not considered in the governing equation.TRM is a numerical approach used to treat latent heat associated with freezing and thawing of water. In this method, temperature analysis is performed at each time step without considering the release of latent heat. If the temperature goes below the freezing point at a given time step, the temperature is brought back to the freezing point by assuming that the amount of temperature decrease from the freezing point is equal to the amount of latent heat released. The simulation continues until no water remains unfrozen. In another approach, the release of latent heat during the phase change between the liquid and solid phases is directly considered in the governing equation accounting for conduction and convection.In this approach, either the unfrozen water content (thus ice content) is obtained by using the General Clausius-Clapeyron (GCC) equation or the ice content is modeled as an exponential function of temperature in so-called the power model. The finite element method was used to discretize the governing equation to simulate temperature changes in a two-dimensional domain with a size of 2 m by 0.1 m. One end of the domain was kept -20 ℃ while the initial temperature of the domain was 18 ℃.
The comparison between the numerical solution using TRM and the analytical showed that the TRM method achieved high accuracy. The TRM result however showed stair-like fine stagnation during the temperature drop. As the GCC model and the power model revealed a sudden and discontinuous increase in the apparent heat capacity with the onset of freezing due to latent heat, the numerical solutions can be unstable and challenging. Compared with the TRM, both approaches resulted in much slower temperature changes. The impact of latent heat considerably decreases the apparent thermal diffusivity α below 0 ℃, slowing down the temperature decrease and leading to a large difference from the results of the TRM.
TRM was effective for ground freezing analysis and achieved higher accuracy than the methods incorporating latent heat directly into the governing equations. Differences in the modeling of thermophysical properties were found to have a significant impact on the freezing analysis.