Prediction Approach Of Permeability Coefficient Of Unsaturated Soil Based On Nuclear Magnetic Resonance Technique

Permeability coefficient reflects the flow of water in the pores of soil and the seepage characteristics of soil.In addition,it is basic parameter of mechanics characteristic of soil.In engineering application,permeability coefficient has great influence on the structural stability,which mainly reflects on some engineering and daily problems,such as the influence of rainfall on the slope stability analysis,subgrade deformation,leakage problem of flood prevention embankment and clay wall of earth-rockfill dam,the influence of landfill and nuclear waste landfill on the pollution and diffusion of groundwater,the migration of underground sewage and so on.However,it is not very realistic to measure the permeability coefficient of unsaturated soil by experimental method due to the complex and time-consuming experimental process.Therefore,it is very important to investigate the indirect prediction method of unsaturated permeability coefficient.Taking Hunan clay as the research object,soil-water characteristic curve test,saturation / unsaturated permeability test and nuclear magnetic resonance(NMR)test are carried out,respectively.Based on the analysis of experimental data and microscopic pore channel theory,a method of predicting unsaturated permeability coefficient from NMR test is proposed,and then compared with the method of predicting unsaturated permeability coefficient from soil-water characteristic curve.The main research contents are as follows:(1)The soil-water characteristic curve of samples with different dry densities is measured by pressure plate apparatus.To obtain the continuous soil-water characteristic curve,Fredlund-Xing model is selected to fit the discrete test data points.(2)Because the test soil sample is clay,the saturated permeability coefficient of samples with different dry densities is measured by variable head test device,then the saturated permeability coefficient at room temperature is obtained.Meanwhile,the unsaturated permeability coefficient of samples with different dry densities is measured by plexiglass bucket device.Through the analysis of experimental data,the variation of unsaturated permeability coefficient with moisture content and matrix suction is obtained.(3)NMR tests of samples with different dry densities during drying and wetting process are carried out by using low magnetic field NMR instrument developed by Suzhou Niumag Company.Based on the microscopic pore channel theory,the signal amplitude corresponding to the transverse relaxation time of each stage is converted to the corresponding volumetric water content,then the unsaturated permeability coefficient can be predicted.In addition,a optimal NMR prediction method is selected by comparing the prediction results of drying and wetting NMR tests with that of one-time saturated NMR test.(4)Mualem model is selected to predict unsaturated permeability coefficient.Then,the prediction results of soil-water characteristic curve method and the optimal NMR prediction method are respectively compared with the measured values to verify the reliability of the proposed model.