Microstructure Research On Spatial Variability Of Clay In Hefei District

Spatial variability of soil parameters is the main reason for the uncertainty of geotechnical engineering,and it's an inherent characteristic of soil.And research on soil microstructure has a great significance on theoretical exploration and engineering practice.Thus,spatial variability analysis based on microstructure was conducted to estimate the auto-correlation length more efficiently.In this paper,9 kinds of unsaturated clay in Hefei were studied.Laser method was used to test the particle-size distribution curve of every sample point.The difference of particle-size distribution betweem different sample point was compared.In addition,the compositions of tested soils were classified in order to obtain the basic particle size composition information of testes soils.The image of soil mocrostructure was obtained by scanning electron microscope(SEM)test.Matlab was used to the image of SEM for quantitative analysis of soil information that including soil particles and voids.Besides,a SEM-based method for estimating parameters that used in construct microstructure by the quartet structure generation set(QSGS)was proposed.According to soil microstructure that constructed by QSGS,the n-point spatial correlation function that based on the theory of Monte Carlo simulation was applied for calculating the value of the auto-correlation length by using correlation function method.The n-point spatial correlation function measures the probability of finding n points located on the space occupied by the same phase of a heterogeneous material.Besides,the value of the auto-correlation length will probably be influenced by mesh density and model size.Therefore,eight groups of different mesh densities and eight groups of model sizes are analyzed,respectively.Based on the similarity theory,the paper used dimensional analysis to obtain the value of soil parameters in actual research field,and then the auto-correlation length of research field was obtained.The study provided a reference for the study of soil spatial variability by numerical analysis.