Tetsunari MORIYAMA

3D Numerical Simulation on Moisture-heat Coupling Behavior of Artificial Ground Freezing Protection in Tunneling

Mitsutaka SUGIMOTO

One of the methods to create a stable working environment around tunnel and seal the tunnel periphery against underground water is artificial ground freezing (AGF) technique. The AGF technique consists of letting a coolant fluid, with a temperature lower than the surrounding ground, circulated inside probes positioned along the perimeter of the gallery. Currently, there are two ground freezing systems being used in geotechnical applications. One is the liquid brine system and the other is the liquid nitrogen system. Ground initially freezes around the individual freeze pipes by forming individual frozen ground columns around the freeze pipes. With continuous freezing, these individual frozen columns merge as a frozen ground mass and become the freezing curtain.
The formation of freezing curtain during the AGF process is crucial as it serves to control groundwater flow and enhances the strength of the soil. However, obtaining accurate predictions of the freezing curtain remains difficult. This can be attributed to the monitoring limitations that hinder the collection of field observations. In addition, the network of freezing tubes is generally designed to be inclined radially; therefore, a three-dimensional (3D) model should be established because of the different arrangements and positions of freezing tubes in each axial section of the tunnel.
To address these limitations, this study presents a simple moisture-heat coupling model that combines heat transfer mechanisms of ice/water phase transition, and the Darcy equation of porous media to predict the dynamic formation of the freezing curtain in a 3D model. A case study of the development of a freezing curtain around Hazawa tunnel construction during horizontal AGF is presented. To confirm the feasibility of the model, temperature data were continuously collected from the case study for more than 35 days. The proposed model delivers temperature results that are consistent with the field data.