冻土水热特性测定与模拟：热脉冲—时域反射技术应用

发布时间：2018-03-09 20:40

本文选题：Thermo-TDR　切入点：冻土热特性　出处：《中国农业大学》2016年博士论文　论文类型：学位论文

【摘要】：在冻土的水热耦合迁移研究中,冻土热特性和冻结特征曲线的测定与模拟以及冰含量的测定具有非常重要的意义。传统的稳态法在测定冻土热特性时经常存在一定误差,热脉冲方法则在测定过程中可能引起冻土中冰的融化；热特性预测模型广泛应用于土壤水热耦合传输模拟中,但是目前缺乏一个简单、准确、未冻土和冻土都适用的热导率预测模型。冻土中冰含量还没有一个标准测定方法,已有的方法都有一定的限制。冻结特征曲线是冻土研究中重要的水力学参数,但其测定过程常常受到过冷却现象等影响。热脉冲-时域反射技术(T hermo-TDR)可以同时测定土壤的液态水含量和热特性,通过改进加热方式能将其应用到冻土中。另外Thermo-TDR具有间接估计冻土冰含量的潜能,也可以用于冻结特征曲线的测定。本研究旨在通过优化TTiermo-TDR的加热方式,使其能够准确测定冻土的热特性,并利用热特性数据改进热导率模型和估计冻融过程中土壤冰含量变化。此外,本研究还改进了冻结特征曲线的测定方法,探讨利用冻结特征曲线来预测其他土壤物理特性。主要研究结果如下：第一,对冻土中Thermo-TDR加热模式进行了优化,使其能够准确测定冻土的热特性。当土壤温度低于-5℃时,采用加热时间60s和加热量450J m-1或者加热时间90 s和加热量450-600 Jm-1的加热模式,可以有效低限制冻土中冰的融化,得到准确的热容量和热导率。第二,发展了一个简单、准确且同时适用于未冻土和冻土的热导率预测模型。该模型以deVries(1963)模型为基础,通过简化和优化原模型的一些参数,能够准确地预测未冻土和冻土的热导率。利用实测和文献数据对简化de Vries模型准确性的检验表明,相比于已有模型,简化模型在未冻土和冻土中预测结果的均方根误差(RMSE)都最小。第三,可以利用TTiermo-TDR测定的冻土的容积热容量、液态水含量以及土壤容重来间接估计土壤冰含量。利用TTiermo-TDR测定了不同初始含水量的3种不同质地土壤冻融过程中的冰含量变化。结果表明,当温度低于-5℃时,TTiermo-TDR测定的冰含量误差主要在±0.05 m3m-3范围内,但在质地较粘土壤中测定误差比较大(达到±0.10 m3 m-3)。第四,可以利用冻土热导率结果和简化的de Vries模型来反推得到冰含量。敏感性分析表明,热容量法比热导率法对于液态水含量等误差源更敏感,热导率法估计的冰含量要比热容量法的结果更准确。在3种土壤上冰含量的实验结果证实了这一理论推断。第五,利用冰牙签法消除了过冷却影响,并进一步了比较平衡法与采用不同降温速率的动态法测定的冻结特征曲线。结果表明,与平衡法相比,采用10℃ h-1降温速率的动态法的测定结果存在一定的误差,而采用2.5℃ h-1降温速率的结果与平衡法没有差异,动态法实验操作更简单。第六,对不同冻结特征曲线拟合模型准确性的比较指出,分段指数函数模型拟合效果最好。联合冻结特征曲线和沙箱法数据拟合了5种不同质地土壤的全范围水分特征曲线。结果表明,除了砂壤土外,其他四种土壤的拟合曲线与沙箱法、压力板法和露点水势仪法实测的全范围含水量-水势数据具有很好的相关性,相关系数在0.948和0.985之间。冻结特征曲线上-10℃以下的数据可以用来估计土壤的比表面积,估计结果与露点水势仪测定值相关系数为0.985。
[Abstract]:In the frozen soil water heat coupling migration study, has a very important significance to determination of thermal properties of frozen soil and determination of freezing characteristic curve and simulation and ice content. The conventional steady-state method often has some error in the determination of thermal properties of frozen soil, heat pulse method may cause the permafrost melting ice in the determination process of heat; the characteristic model is widely used in the coupled water and heat transfer simulation of soil, but the lack of a simple, accurate, predictive model of thermal conductivity without permafrost and the permafrost applicable rate. In the frozen soil ice content determination method is not a standard, the existing methods have certain limitations. Freezing characteristic curve is an important hydraulic study on parameters of frozen soil, but the determination process is often influenced by cooling phenomenon. Thermo time domain reflectometry (T hermo-TDR) liquid water content determination of soil and thermal characteristics, By improving the heating method can be applied to the soil. In addition Thermo-TDR has indirect estimation of permafrost ice content potential, can also be used for the determination of freezing characteristic curve. The purpose of this study is to optimize the TTiermo-TDR through heating, the thermal characteristics of accurate measurement of frozen soil, and the thermal characteristic data of ice content changes of soil thermal conductivity model and the estimation of the freezing thawing process improvement. In addition, this study also improved the determination method of freezing characteristic curve, to explore the use of freezing characteristic curve to predict the physical properties of other soil. The main results are as follows: first, the Thermo-TDR heating mode in frozen soil were optimized. The thermal characteristics can be accurately determined. When the soil is frozen the temperature is lower than -5 DEG C, the heating time and heating capacity of 450J 60s M-1 90 s or heating time and heating rate of 450-600 Jm-1 heating mode, can effectively lower limit Frozen in the ice melts, accurate heat capacity and thermal conductivity. Second, the development of a simple, accurate prediction model of thermal conductivity and also applicable to frozen soil and permafrost. In this model, deVries (1963) model as the foundation, through the simplification and optimization of some parameters of the original model, can not accurately predict the frozen soil and permafrost thermal conductivity rate. Using the measured data to test and document the simplified de Vries model accuracy show that compared to the existing simplified model, the root mean square error of prediction results in frozen soil and frozen soils (RMSE) are minimum. Third, the volumetric heat capacity determined by TTiermo-TDR liquid water content of frozen soil. And the capacity of soil heavy soil ice content. Indirect estimates of 3 different kinds of soil freezing and thawing process of the ice content changes of different initial water content was determined by TTiermo-TDR. The results show that when the temperature is lower than -5 DEG C, TTier The ice content error measured by mo-TDR mainly in the 0.05 m3m-3 range, but the texture is sticky soil determination error is relatively large (up to 0.10 m3 M-3). De fourth, Vries model can be simplified by using frozen rate results and the thermal conductivity of ice to determine the content. The sensitivity analysis showed that the thermal conductivity method specific heat capacity method more sensitive to the error of liquid water content source, the amount of ice thermal conductivity estimation to specific heat capacity and the method is more accurate. In 3 kinds of soil ice content was confirmed by the experimental results that this theory. Fifth, use ice toothpick method to eliminate the cooling effect, and further compared with equilibrium method dynamic method at different cooling rates for the determination of freezing characteristic curve. The results show that, compared with the results of the dynamic balance method, determination method with 10 DEG H-1 cooling rate has some error, and using H-1 2.5 degrees of cooling rate and results There is no difference between the method of dynamic balance method, the experimental operation is simple. Sixth, comparing to different freezing characteristic curve fitting accuracy of the model, piecewise exponential model best fitting effect. Combined with freezing characteristic curve and the sandbox method of data fitting 5 different soil texture full range water characteristic curve. The results show that, in addition to outside sandy loam the other four kinds of soil, the fitting curve and the sandbox method, the full range of pressure plate and dewpoint waterpotential instrument method to measure the water content and water potential data has good correlation, the correlation coefficient between 0.948 and 0.985 degrees Celsius. Frozen data -10 characteristic curve can be used to estimate the soil surface area, the estimated value of the relevant coefficient of determination results and dewpoint waterpotential instrument 0.985.

【学位授予单位】：中国农业大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：S152

【相似文献】