Reliability Analysis Of Slope Stability And Rainfall Landslide Risk Analysis Based On Random Field Theory

Landslides are among the world's most frequent geohazards and pose serious risks to human activities on slopes across the globe.In recent years,global warming has received widespread attention.Abnormal extreme weather and heavy rains often cause landslides and debris flows,resulting in huge loss of life and property.Therefore,it is of great significance to study the risk of landslide under rainfall conditions.In addition,the variability of soil parameters is an important issue of uncertainty in geotechnical engineering.The traditional slope stability analysis adopts deterministic analysis method,and the safety factor is used as the evaluation indicator of the slope stability,which can lead to unconservative estimationresults.Therefore,considering the spatial variability of soil parameters,the slope stability and reliability are analyzed.It is more reasonable to evaluate the slope stability by combining the safety factor and the failure probability.In this thesis,the Karhunen-Loeve method is used to generate the random field of soil parameters.The flowchart of reliability analysis of slope stability based on GeoStudio and OptumG2 software packages using a batch mode is illustrated.The slope reliability problems and the rainfall-induced landslide risk,considering the multiple spatially correlated soil properties are studied.Based on the strength reduction finite element limit analysis,the reliability analysis framework of slope stability is established,and the results are compared with the traditional limit equilibrium methods.Furthermore,the influence of variability of soil strength parameters and vertical correlation length on slope stability were studied.In addition,based on the random field theory and the unsaturated seepage theory,the process of rainfall infiltration in a specific slope is simulated by random seepage analysis.The influences of the spatial variability of the soil saturated permeability coefficient and the time variability of rainfall on the stability of slope were studied.And the risk of rainfall landslide is evaluated by the landslide mass as the criterion of failure consequence.The main conclusions of this thesis are as follows:(1)A variety of failure modes of the slope coexists when considering the spatial variability of soil parameters,which is significantly different from the single failure mode obtained by deterministic analysis.The finite element limit analysis can effectively search for a series of possible critical slip surfaces under the condition of spatial variation of parameters,and the strict upper and lower limits of the slope safety factor are determined.For the simple homogeneous slope,the results given by finite element limit analysis and limit equilibrium methods are basically consistent.When the spatial variabilities of soil parameters are large,the limit equilibrium method can not give reasonable modes of landslide.The variability of soil strength parameters leads to significant changes in the safety factor probabilistic distribution.(2)Considering the spatial variability of soil parameters,the slope safety factor decreases with the increase of rainfall intensity.When the rainfall intensity equals to the the critical rainfall intensity,i.e.the mean value of the permeability coefficient,the safety factor of the slope is the lowest.Henceforth,the increase of rainfall intensity no longer causes changes of the slope safety factor.(3)Under the critical rainfall intensity,and with the increase of the variability of the saturated permeability coefficient,the mean value of the slope safety factor increases gradually,while the failure probability and the landslide risk are gradually reduced.Even with the same slope failure probability,the corresponding landslide risk indicators are different.The landslide risk index accounts for the landslide characteristics more comprehensively and can better assess the failure mode quantitatively.In the rainfall infiltration stage,the greater the vertical correlation length,the easier the rainwater infiltrates into the slope,so the probability of deep-seated landslide gradually increases.With the increase of the vertical correlation length,the failure probability and the landslide risk gradually increase.This thesis has done the following innovative work:(1)Slope stability analysis considering spatial variability of soil parameters,unlike the traditional limit equilibrium method and finite element method,the finite element limit analysis does not need to assume the slip surface in advance,and overcomes the defects that the failure criterion is not unique.Therefore,based on this method,the slope stability reliability analysis framework is established,this method can effectively search for a series of possible critical slip surfaces under the condition of spatial variation of soil parameters,and the strict upper and lower limits of safety factor are obtained.(2)The current slope stability analysis methods neglect the effects of rainfall intensity,duration and other temporal/spatial variabilities of saturated permeability coefficient on landslide.Therefore,based on the random field theory and the unsaturated seepage theory,the stochastic seepage analysesare carried out to study the effects of rainfall intensity and spatial variability of saturation permeability coefficient on landslide risk.A more practical method for the reliability of slope stability is then established,which considering the time-space composite random field of soil parameters under rainfall conditions.