Variability Analysis Of Undrained Shear Strength Of Coastal Soft Soil

Soft soil is widely distributed in coastal areas of China,and its engineering characteristics is affected by different sedimentary environments such as river alluvial and transgression.Undrained shear strength is an important parameter for evaluating the bearing capacity of foundations.Whether it is the selection of parameter characteristic values in deterministic design or reliability theory,the variability of geotechnical engineering parameters is to necessary for soft soils.With the support of the National Key Program,the sedimentary history and spatial distribution of the soft soil in coastal areas of China is carried out;In the absence of shear test data,the Bayesian theory is introduced to optimize the posterior distribution of undrained shear strength;Based on the correlation between the cone tip resistance in CPT test data and the undrained shear strength,a random field theory is introduced to characterize the spatial variability of undrained shear strength,and the spatial characteristics of soft soils with different geological causes are studied;Combined with the random theory,the reliability level of the foundation bearing capacity is learned.The main content and results are as followed:（1）Extensive researches are carried on the regional geology,borehole data and regional specifications of coastal areas in China.The results show that coastal areas such as Hebei,Jiangsu,and Zhejiang suffered four to five transgressions,and only one transgressive layer remained in southern China;the sedimentary facies of the soft soil can be divided into littoral facies along the coastal area;the delta facies in the Yellow River Flooded Plain,the Yangtze River in northern Jiangsu,the lower reaches of the Qiantang River and the lower reaches of the Pearl River;the lacustrine facies in the Taihu Lake;the lagoonal facies in the Lixia River area of Jiangsu;the drowning facies in the Minjiang River in Fujian;The thickness of the soft soil can be up to about 40m,and the thickness shows a trend of decreasing from the coast to the mountain.（2）When only limite shear test data is available on the site,the Bayesian theory is introduced to use the empirical distribution of undrained shear strength to reduce the variability of undrained shear strength.The preconsolidation pressure has a good relationship with the liquidity index.By comparison,σ’_p/p_a=10^0.9-0.96LIis more suitable for soft soils with a liquidity index more than 1;Without the direct shear test,Bayesian theory is introduced to use several sets of liquidity index test values and a prior information of undrained shear strength,based on the above formula,establish the probability density function of the posterior distribution,use the Monte Carlo method to solve by generating equivalent sample of drainage shear strength,and it can be used in small and medium-sized projects.（3）The undrained shear strength data are continuously generated based on the cone tip resistance in the CPT test,and random field theory is introduced to calculate the spatial variability of the undrained shear strength;Based on the collection of a large number of CPT test data,the spatial characteristics of the silty clay layer of the sea-land interaction and river deposition are studied.The results show that the correlation distances of the soft soils are different due to the differences in the sedimentary environment;the soft soils deposited by the land and sea alternately are easy to form the particle size distribution due to the impact of the sea regression rhythm layer and thin sand layer,the parameter value is more discrete,resulting in the correlation distance is concentrated in the low value interval.The correlation distance ratios calculated from the cone tip resistance and the lateral friction resistance are both 0.91.The correlation distance parameters calculated from the two parameters are similar,which is consistent with the concept that the correlation distance is an inherent property reflecting the correlation characteristics of the soil.（4）Considering the spatial variability of undrained shear strength,random field theory is introduced to study the reliability analysis of foundation bearing capacity with CPT test.If the parameter spatial variability is considered,the reliability index and failure probability calculation results are consistent with sufficient safety factor reserves;On the contrary,the traditional approach makes the parameter variability too large,so that the reliability index is too low and the failure probability is too high,which is not consistent with the actual situation.Compared with the weight of the soil,the variability of undrained shear strength has a greater impact on the reliability analysis of the bearing capacity;the value of the relevant distance is also more sensitive to the reliability analysis results.