| In cold regions,the ground is subjected to freezing process because of the decrease of temperature.Freezing process changes the hydro-thermal characteristics of soils and produces water-heat-vapor transfer,which results in frost heave and induces thaw settlement during the thawing process.A great number of engineering diseases occur because of frost heave and thaw settlement,which severely affect long term performance of cold regions engineering and its service life.Therefore,it is necessary to completely understand the hydro-thermal characteristics and water-heat-vapor mechanisms of soils during the freezing process,which will benefit for the repair of the existing cold regions engineering and prevention of engineering diseases in the future cold regions engineering.In this study,indoor tests,theoretical analysis,and numerical modelling are introduced to analyze these scientific problems.The main research contents and conclusions are as follows:(1)Many unfrozen water content models and selected six widely used unfrozen water content models were analyzed in this study.The effect of parameters on the unfrozen water content models and sensitivity of unfrozen water content models to the measured subzero temperature ranges were investigated.The Blanchard and Frémond model,McKenzie et al.model,and Kozlowski model contain a critical unfrozen water content,suggesting that further decrease of subzero temperature does not affect the calculated volumetric unfrozen water content when the temperature is below a critical temperature.The FX-Clapeyron model and FX(C(T)=1)-Clapeyron model do not have a critical unfrozen water content and the calculated volumetric unfrozen water contents decrease with the decrease of the subzero temperature.The VG-Clapeyron model has the characteristics of above models in the range of-50°C and 0.The Kozlowski model and FX-Clapeyron model are the least sensitive to measured subzero temperature ranges.The calculated unfrozen water contents determined by the FX-Clapeyron model usually perform best compared to those determined by the other five models,especially for smaller subzero temperature ranges.(2)The effects of initial water content,dry density and soil type on the freezing and thawing temperatures during a freezing-thawing process were investigated.The initial water content and soil type have significant effects on the freezing and thawing temperatures,while the dry density has an insignificant effect on the freezing and thawing temperatures.Based on the unfrozen water content model,a new mathematical formula was proposed to calculate the freezing and thawing temperatures.A total number of 29 soils(different soil types,different soil properties and soils with different salt contents)were used to verify its effectiveness in calculating the freezing and thawing temperatures.Results show that the calculated results of the new model are in good agreement with that of the Azmatch model,suggesting that the new model can be used to determine the freezing and thawing temperatures of various soils with a wide range of soil properties.(3)For the sands with a high amount of pure quartz contents,a new thermal conductivity model at dried condition was proposed.A comparison of the new model and 3 models shows that the new model performs best among the 4 models to calculate thermal conductivity of dried sands with a high amount of pure quartz contents.From analyzing the applicability and limitation of the thermal conductivity models in the literatures,a thermal conductivity model for fine-grained soils and a thermal conductivity model for coarse-grained soils at the unfrozen state were developed using the Johansen's method.30 fine-grained soils and 29 coarse-grained soils were selected from the literatures to develop a thermal conductivity model for fine-grained soils and a thermal conductivity model for coarse-grained soils,respectively.6 fine-grained soils were used to evaluate the thermal conductivity model for fine-grained soils and 20 coarse-grained soils were used to evaluate the thermal conductivity model for coarse-grained soils.Results show good agreements(R~2>0.89)between calculated and measured thermal conductivity,indicating that the two models have good performance in calculating the thermal conductivity of unfrozen soils.(4)The freezing process of soils are divided into three stages by soil components and frost heave:(1)Stage 1,a soil sample does not have frost heave and contains four components,e.g.unfrozen water,ice,solid grain and air;(2)Stage 2,a soil sample does not have frost heave and contains three components,e.g.unfrozen water,ice and solid grain;(3)Stage 3,a soil sample has frost heave and contains three components,e.g.unfrozen water,ice and solid grain.On the basis of the parallel-series model for unfrozen soils and different connections(e.g.series connection or parallel connection)between unfrozen water and ice in the pores,a generalized model was developed in each stage.The model was verified by the cited measured data(8 silty clay samples and 3 aeolian sand samples)from previous articles.Results show good agreements(R~2>0.91)between calculated and measured thermal conductivity,indicating that the new model has good performance in calculating the thermal conductivity of freezing soils.(5)The effects of initial water content of coarse-grained soils,initial water content of fine-grained soils,water supply condition,cold end temperature,warm end temperature and position of the second plastic wrap on the water-heat-vapor transfer were investigated.These factors significantly affect the water redistribution and result in different numbers of peak water content.A thin ice layer forms at the top surface of soil sample when the freezing test is finished.It increases water content at the top surface of soil sample and verifies the water-heat-vapor transfer phenomenon.A higher initial water content of coarse-grained soils tends to result in a higher final water content at the top surface and freezing front of soil sample.A higher initial water content of fine-grained soils tends to result in a higher final water content at the top surface of soil sample,but the peak water content only appears at the top surface of soil sample.Water supply boundary slightly affects the final water content at the top surface of soil sample and has a great influence on the final water content at the freezing front and unfrozen zone.A lower cold end temperature and a higher warm end temperature increase the temperature gradient,which promotes vapor and liquid water transferring to the colder end and increases the final water content at the top surface of soil sample and freezing front.When the second plastic wrap is above the freezing front,there are three peak water contents in the soil sample at the end of freezing test.The first peak water content is at the top surface of soil sample,the second peak water content is below the second plastic wrap and the third peak water content is at the freezing front.When the second plastic wrap is below the freezing front,there are two peak water contents in the soil sample at the end of freezing test.The first peak water content is at the top surface of soil sample and the second peak water content is at the freezing front.When the second plastic wrap is above the freezing front,water supply condition slightly affects the final water content at the frozen zone and has a significant influence on the final water content at the unfrozen zone.(6)On the basis of governing equations for mass conservation and energy balance during the freezing process,Comsol was used to simulate the freezing processes of two soil samples.One soil sample has a plastic wrap and it locates at the top surface of soil sample.The other soil sample has two plastic wraps,the first plastic wrap locates at the top surface of soil sample and the second plastic wrap is above the freezing front.Results show that the simulated temperature and water fields are in good agreements with the measured results,which illustrates that the model can effectively simulate the water-heat-vapor transfer for coarse-grained soils. |
PDF Full Text Request