Preliminary Study On Application System Of Random Field Method In Geotechnical Engineering Stability And Deformation Problems

Due to geological effects,natural geotechnical soil has significant spatial uncertainty,and the impact on large-scale engineering cannot be ignored.The research on geotechnical engineering uncertainty is in the ascendant,but it is difficult to realize engineering application due to the limits of survey technology,experimental methods and analytical methods.The paper attempts to establish an application system of the random field method based on data processing,random field analysis and reliable evaluation of geotechnical engineering stability and deformation problems,and demonstrates in the actual engineering of flysch slope.In the data processing,the geotechnical data is divided into characteristic data,prior statistical data and posterior statistical data.The stiffness,strength and permeability parameters in the Mohr-Coulomb model are taken as key variables,and the random field database is established based on its relationship with geotechnical tests and hydrogeological test indicators.The database retains the feature point parameters,applies the variogram theory to obtain the related distance parameters and reconstructs the twodimensional and three-dimensional random field models based on the Karhunen-Loeve expansion in numerical experiments.Furthermore,the influence of the posterior statistics is described according to the Bayesian theory.Random field analysis includes stability and deformation.The random field based on finite element limit analysis and Monte-Carlo simulation are used to study the stability problem.The two-dimensional feature random field is established to study the layered slope,foundation pit safety,support structure efficiency and consider the seepage and rainfall effects.The stochastic column is used as the three-dimensional random field simulation unit to analyze the stability of group piles and linear structures.The uncertainty analysis can reflect the extreme failure mode,and the safety factor obeys the normal distribution and is sensitive to the coefficient of variation of the key variables.The coefficient of variation increases,the concentrated distribution of weak areas may increase,the range of sliding damage expands and the supporting structure needs to provide greater resistance.Therefore,the mean value of safety factor decreases,the sliding volume,the mean value of internal force of supporting structure increases,the standard deviation increases.The coefficient of variation of key variable above the threshold requires consideration of the potential risks implied by uncertainty.The random field based on finite element elastoplastic analysis and Monte Carlo simulation are used to study the deformation problem.The preset data monitoring points extract the global displacement,local displacement and structural internal force index.The twodimensional problem studies the deformation of the layered slope and the restraining efficiency of the supporting structure,as well as the deformation of the pit and the disturbance to the surrounding building structures caused by the excavation of the foundation pit.The variability of the simulation results is positively correlated with the spatial variability of the stiffness parameters,but much smaller,because the support structure can effectively restrain extreme deformation.The combination of the anchor and the retaining wall has the best control deformation effect,and the monitoring point frequency histogram can increase the structural design and deformation evaluation accuracy.The difference of settlement,rotation angle and roundness deviation ratio obtained by the deterministic analysis of excavation disturbance is low,and the risk is underestimated.The random block is used as a three-dimensional random field simulation unit to simulate group piles and linear structures,and the application range of random field deformation analysis is extended.In the reliable evaluation,the statistical methods of the simulation results are combed,and the data are compared by means of the probabilistic indicator tool.The forward or reverse direct evaluation method is used to check the reference value of the direct evaluation method of the engineering result index.When it is characterized as a normal distribution function,Multiple sets of inspection indicators ? can be integrated to further give reliable indicators of geotechnical engineering systems.