Research On The Spatial Variability And Robust Geotechnical Design Of Excavation In Soft Soils Based On Cptu Data

With the national economy growing,urban construction in China is so fast,and there are more and more underground projects,especially in coastal cities.Most deep excavation projects are located in soft soil,which puts forward more strict safety requirements.Soils has been formed in a long geological history,thus remarkable Spatial variability occurs when geotechnical parameters of soils that are measured at different spatial locations exhibits values that differ across the locations.The complex composition and structure of soft soils result in a high borehole sampling disturbance and low reliability of labs experiment results.To evaluate the spatial variability of geotechnical parameters of soft soils,multifunctional piezocone penetration test(CPTU)is in urgent need.The spatial variability of geotechnical parameters leads to uncertainty in geotechnical design(eg.uncertainty in system response).In order to reduce the sensitivity of the system response of geotechnical design to the spatial variability of the geotechnical parameters,the geotechnical robust design method puts to use.Robust geotechnical design seeks the reduction in the variation of system responses by adjusting only the“easy-to-control”design parameters,and not the“hard-to-control”noise factors,on the basis of uncertainty analysis of geotechnical parameters,simultaneously satisfying safety and cost requirements.The paper is supported by the National Key R&D Program of China,the National Key Research and Development Program Subproject,the National Natural Science Foundation Project,the Foundation of Jiangsu Province Outstanding Youth and the Jiangsu Provincial University Graduate Research Innovation Program.The discerning results of homogeneous soil units(HSU)of soft soils are presnented,and the spatial variability of of soft soils are studied detailed based on the identification results of homogeneous soil uinsts with CPTU data.A conditional simulation method of non-stationary log-normal distribution random field is proposed to study the spatial distribution of soil parameters with a depth trend and anisotropy in scale of fluctuation.The Monte Carlo simulation was used to analyze the effects of spatial variability of undrained shear strength on the reliability of excavation in soft soil.The Robust geotechnical design of braced excavation is achieved by optimizing the system response and design cost considering the spatial variability of geotechnical parameters.Here are some research conclusions:(1)Identification of homogeneous soil units based on CPTU.The undrained shear strength is predicted based on the CPTU data,match the results of Field Vane test very well.In the soil layer with a same soil structure and CPTU borehole,the identification results of homogeneous soil units of different soil parameters differs from each other,and there is no correlations between them.The homogeneous soil uints distributed closely in the Robertson soil classification chart,and the variability of soil parameters is relatively low.(2)Spatial variability of homogeneous soil unit parameters based on CPTU data The spatial variability of soil parameters of homogeneous soil units is studied detailed with the random field model.In the vertical direction,a rigorous method is used to study the spatial variability of soil parameters,the modified Bartlett test is used to inspect the stationary of fluctuation components.In the horizontal direction,the expeditive method is used to estimate the spatial variability of soil parameters.The results show that there is no specific autocorrelation function model in the site describing the random field of marine soft soil well.The spatial variability of cone tip resistance and undrained shear strength has remarkable depth trend and anisotropy in scale of fuctuation.Generally,the fluctuation range in the horizontal direction is an order of magnitude higher than the fluctuation range in the vertical direction.(3)A conditional simulation method for the non-stationary log-normal random field of soil parameters is proposed.Combined with the CPTU data,the conditional probability density of the spatial distribution of soil parameters is derived.By using the conditional simulation method,the spatial distribution of soil parameters with a depth trend and aniostropy in scale of fluctuaiton is simulated.The method uses the LDL~T method to decompose the conditional covariance matrix.There are no requirements for the trend function and covariance structure of the soil parameters.It is also suitable for the conditional simulation of non-stationary lognormal random fields with arbitrary grid.(4)Revealing the influence of spatial variability of undrained shear strength on the reliability of braced excavation in soft soil.The impact of the coefficient of variation,scale of fluctuation and spatial reduction of undrained shear strength on the reliability of braced excavation in soft soil are studed with conditional simulation method.The results show that the probability density function of the safety factor of the basal heave failure obeys the lognormal distribution.The probability density function of the safety factor for push in failure is difficult to estimate.If the influence of scale of fluctuation is neglected,the failure probability will be significantly overestimated.The vertical scale of fluctuation affects the failure probability much greater than the horizontal scale of fluctuation.Smaller scale of fluctuation will cause more spatial averaging effects,lowering the coefficient of variation of soil parameters,and decreasing the failure probability.(5)A robust design method for braced excavation is proposed considering spatial variability of soil parameters.For the braced excavation design,the spatial variability of the undrained shear strength is introduced into the robustness analysis of system response of the geotechnical design.Taking into account of system response and construction cost,multi-objective optimization method is used to achieve robust design results.The comparison results show that the system function response of the robust geotechnical design is significantly smaller than the change of the system function response of the original design scheme.The maximum internal force of the diaphragm wall of the robust geotechnical design scheme is only about 60%of the original geotechnical design.The cost of robust geotechnical design is significantly less than the cost of original geotechnical design.The design area of rebar of robust geotechnical design is only about 70%of the actual design.