Abstract
Artificial ground freezing is a frequently used method for temporary support of the soil in geotechnical interventions, such as tunneling. Besides the time required to form a supporting frozen soil structure during freezing, the prediction of temperature evolution during the thawing process and the final thawing time are also relevant, since settlements induced by thawing of the frozen soil can pose a risk to sensitive structures in urban tunneling. The freezing and thawing process of soils depends strongly on the unfrozen liquid content. The unfrozen liquid content during the freezing and thawing cycles is known to have a pronounced hysteresis response, whose effect on the temperature distribution and groundwater flow, especially at the structural level has not been extensively studied. In this work, we present a thermo-hydraulic finite element model for freezing soils and incorporate the hysteresis behavior of soil. We aim to investigate the suitability of the presented numerical model for tunnel construction on the one hand, and quantify the extent of the influence of the hysteresis response of the unfrozen liquid content at the structural level on the other. To this end, the model is first validated with the help of experimental data from a soil test subjected to a freezing–thawing cycle and subsequently used to analyze three problems of artificial ground freezing in tunnel construction during the freezing phase or freezing–thawing phases. The results show a good agreement with the experimental measurements. We show that the computational model is capable of providing accurate prognosis of the temperature profile as well as broader metrics such as the thickness and shape of the frozen body for tunneling applications under hydrostatic and seepage flow conditions. The hysteresis response is shown to have a significant influence on the coupled thermo-hydraulic process, with thawing times differing depending on whether hysteresis is considered. It is concluded, that the inclusion of the hysteresis response is crucial for obtaining an accurate representation of the freezing process.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have