Unsaturated Hydraulic Conductivity Of A Random Fracture Based On Random Field Theory

Expansive soils have found in more than20provinces in China, which aremainly distributed in Henan, Guangdong and Hubei provinces. Expansive soils caneasily fracture because of its special characteristics of well–shrinking. The presenceof cracks can be induced engineering accidents such as slope instability and pitcollapse. The main reason for this is that the existence of fractures would make theengineering susceptible to the infiltration of rain. And then the bearing capacity willreduce. In order to provide a good basis for engineering design and geologicalhazard assessment, the study of the hydraulic characteristics of fractures is verynecessary.The hydraulic characteristics of fractures mainly include two aspects: waterretention curve and the coefficient of permeability function curve. In this paper, thesaturated-unsaturated coefficient of permeability for a single fracture has beenstudied in depth through theoretical and experimental methods, which have therandom field characteristics.Produce a series of crack aperture data which obey a certain random fieldparameters applying the random field data generating method, and taking intoaccount the trap effect and the capillary law, the program about numericalsimulation on the degree of saturation under different matrix suction for a singlefracture is developed. And then the water retention curve and coefficient ofpermeability curve can be obtained. Compared with the experimental results, thenumerical result show that the numerical method can reflect the water retentioncurves which obtained by experiment have a greater residual moisture under thecondition of considering the trap effect.The random field theory model of fracture is established using the random fieldparameters of the fracture. And then use this model combined with the principles offluid mechanics to solve the upper and lower bound of saturated-unsaturatedcoefficient of permeability for a random single fracture. And then the relationshipsbetween the coefficient of permeability and random distribution parameters alsohave been studied. Study found that the saturated hydraulic conductivity of singlefracture in both limiting case were most affected by the mean value of the aperture.The matrix suction which corresponds to the permeability coefficient began todecline depends on the maximum aperture which the fracture can be achieved. Whenthe discrete element size becomes bigger, the slope of the decline of the hydraulicconductivity becomes bigger.A new capillary barrier was designed, which has a high permeability and a certain air entry value. After experiment verification, the saturated coefficient ofpermeability of the capillary barrier is1.80×10-3m/s; and the air entry value is0.55kPa.Then a device for measuring the unsaturated coefficient of permeability isdeveloped for a single fracture. The saturated-unsaturated coefficient ofpermeability of the single fracture is measured by using this device. Theexperimental results show that as the matrix suction changes, the permeability curvegenerally can be divided into three stages: stable period, fell sharply period andslowly fell to zero. Through the comparison of experimental results and theoreticalanalysis, the results revealed that theoretical analysis results depict the change trendand scope of the coefficient of permeability well. And through the comparison ofexperimental results and numerical analysis result, the results revealed thatnumerical analysis can depict the saturated hydraulic conductivity and the matrixsuction which corresponding to the hydraulic conductivity begin to decline well.