| Soil is a material that is frequently encountered in modern engineering.In natural environment,it may freeze and thaw when the temperature varies near the soil freezing point.Due to multiple reasons,the soil contains unfrozen water even at temperatures below soil freezing point.The research of unfrozen water in frozen soil is very important,since the application lies in many aspects,such as ground source heat pumps,artificial ground freezing technology,freezing and thawing damage in cold regions,and wet porous media.At present,scholars pay much attention to unfrozen water in different types of soils;however,there is no research on unfrozen water in clay minerals,and the analysis of unfrozen water from soil compositions.Therefore,it is necessary to conduct an indepth study on the characteristics of unfrozen water in clay minerals.First,in order to predict the change of unfrozen water in clay minerals,evaluate the prediction effect of the unfrozen water prediction models,and provide a basis for the selection of clay minerals unfrozen water prediction models.The characteristics of unfrozen water with negative temperature during the freezing and thawing of the soil were tested by NMR(Nuclear Magnetic Resonance),and the validity of the latest unfrozen water prediction models were analyzed using the obtained test data.The results showed that,for the prediction of the freezing characteristics of clay minerals,the Adjusted R-square of model Wang and Fredlund reached 0.99,which were recommended to be used first.Model Van Genuchten and Pham reached 0.93,which could be used.Model exponential was only 0.72,which was not recommended.For the prediction of the thawing characteristics,the Adjusted R-square of model Wang,Fredlund,and Pham reached 0.98,which were preferred.Model Van Genuchten and exponential reached 0.94,which could be used.Compared with the freezing process,the effect of model exponential had been significantly improved.For the comprehensive prediction of freeze-thaw characteristics,the Adjusted R-square of model Wang and Fredlund reached 0.99,which were preferred.Model Van Genuchten and Pham reached 0.95 and could be used.Model exponential was only 0.83,which was not recommended.Then,for the NMR test of unfrozen water,whether it is the physical process itself or the test,there will be a certain randomness,so a single unfrozen water test is only a statistical sampling.In order to better evaluate the test results,several tests on the thawing characteristics of montmorillonite were conducted in this paper.Model Wang was used to predict the upper and lower envelope curves.The results showed that both Adjusted R-square reached 0.99,which could accurately reflect the fluctuation range of unfrozen water.The validity of the latest unfrozen water prediction models were analyzed using the overall data,the results showed that Adjusted R-square of the five models for the overall data reached 0.97,and the prediction effect were excellent.The mean value and variance of unfrozen water were analyzed,the results showed that Adjusted R-square of model Wang for the mean value of unfrozen water reached 0.99,and the prediction was accurate.Adjusted R-square of the model polynomial for the variance of unfrozen water reached 0.90,and the prediction was accurate.The distributions of unfrozen water during the test were analyzed,the results showed that unfrozen water test followed a uniform distribution and a normal distribution,but did not follow an exponential distribution.The parameters of each distribution were given and the probability density function images were drawn.Finally,in order to study the influence of clay minerals on the thawing characteristics of soil from the composition,experimental tests on the thawing characteristics of the mixture of three typical clay minerals and sand with different contents were carried out.The validity of the latest unfrozen water prediction models were analyzed using the obtained test data.The results showed that Adjusted R-square of the model Wang and Fredlund for the thawing characteristics of clay minerals sand mixture both reached 0.98,and the prediction effect were good,with model Fredlund being better.The effect of clay minerals on unfrozen water in sand were analyzed.The results showed that unfrozen water content increased with the increase of clay mineral content,and the rate of change were related to the type and content of clay minerals.The mechanism of the influence of clay minerals on the characteristics of unfrozen water of the mixture was revealed,and the strong surface action and small particle size of the clay minerals were responsible for the increase in unfrozen water content as the clay minerals content increased.The unfrozen water prediction models of clay minerals sand mixture were revised to take into account the influence of clay minerals content.Among them,the best model of model parameters changed with clay minerals content were related to clay mineral types.This paper has 61 figures,138 tables,and 79 references. |
PDF Full Text Request