Font Size: a A A

Study Of Rainfall-induced Sandy Loess Landslide Identification,Soil Mechanical Properties Of Slip Zone And Formation Mechanism

Posted on:2023-07-12Degree:DoctorType:Dissertation
Country:ChinaCandidate:Z X LiFull Text:PDF
GTID:1520306845452074Subject:Geological Engineering
Abstract/Summary:PDF Full Text Request
Sandy loess,as a special type of disaster-prone soil,is strongly sensitive to water humidification and excavation disturbance.Under the effect of extreme precipitation and human engineering activities,concentrated and clustered loess landslides occur more frequently in the northern part of Loess Plateau,such as the northern part of Shaanxi Province,where the sandy loess zone is dominant,seriously threatening the safety of people’s lives and properties.It is urgent to study the mechanism of sandy loess landslides to provide a theoretical basis for scientific disaster prevention and mitigation.This thesis is based on the National Natural Science Foundation of China(NSFC)project "Research on the mechanism of artificial vibration for seepage and slide promotion in loess landslides",focusing on sandy loess landslides in the Loess Plateau,and investigating the characteristics of regional sandy loess landslide clusters and disaster identification methods from a multi-scale perspective based on earth system science.The engineering geological investigation has been conducted and a geological model of typical sandy loess landslides has been constructed.Based on the finite element software platform,we have also carried out the numerical simulation of typical rainfall infiltration sandy loess slopes and revealed the formation mechanism of rainfall sandy loess landslides.The following research outcomes have been achieved:(1)Based on the data of landslide hazard register sites and landslide hazard sites occurred in the past three years in the study area of northern Shaanxi,and combined with high-resolution DEM images,we have identified the regional control characteristics of topography and geomorphology,geological structure,hydrogeology,vegetation cover,precipitation activities and human activities and other disaster-preventing factors on sandy loess landslides;we have then constructed a machine learning model of disaster-preventing and triggering evaluation factors,normalized and standardized the multi-source heterogeneous data,and realized the quantitative evaluation of the hazard zone classification of sandy loess landslides in the study area.(2)The adaptive Stacking baseline selection technique has been combined with the regional coherence statistical threshold to screen interferometric pairs to optimize the adaptive temporal baseline;the In SAR-Stacking temporal interferometric processing technique has been used to improve the recognition information capability and recognition efficiency of In SAR technology;furthermore,a set of methods has been proposed for the identification of sandy loess landslides starting from the image morphology and tonal characteristics,and the location,morphology and distribution range of typical sandy loess landslides in the study area are traced.On the basis of this,a more detailed engineering geological survey of the identified typical sandy loess landslides has been conducted,and two types of typical sandy loess landslides,transient and hysteretic,have been proposed and a geological model was constructed.(3)A series of tests on the water-force characteristics of sandy loess in the slip zone with different matrix suction have been conducted to analyze the soil-water characteristics,critical state characteristics and yield characteristics,and to identify the matrix suction correlation of the shear characteristics of sandy loess in the slip zone;the skeletal stress and matrix suction were selected as the intrinsic stress,and an elastic-plastic intrinsic model suitable for describing the water sensitivity of sandy loess in the slip zone has been constructed on the basis of the Barcelona model(BBM),and the intrinsic model was numerically implemented by using an automatic sub-incremental stepwise algorithm with error control.(4)Aiming at the main control inducing factors of hysteretic loess landslides-rainfall and excavation unloading,triaxial creep tests with constant bias stress and axial stress gradually unloading the enclosing pressure path have been carried out on sandy loess in the slip zone with different suction.The characteristics of its strain-time and isochronous stress-strain curve have been analyzed,and the level of impact of unloading stress and its suction on the curve has been identified,and.thus,the long-term strength evolution pattern according to suction has been obtained;then,the yield proximity YAI criterion has been introduced,and the nonlinear viscous pot element has been connected with the Burgers model in series,and a modified Burgers model has been proposed to describe the creep behavior of sandy loess under different unloading stress paths.(5)Based on the UMAT interface of ABAQUS finite element software,a subroutine of the constructed sandy loess in the slip zone intrinsic model has been developed in the FORTAN language;then,numerical simulations of typical sandy loess slopes under rainfall infiltration conditions have been carried out to reveal the evolution of seepage,deformation and stress fields of the slope body under different rainfall intensities and to analyze the occurrence,development and penetration processes of the plastic zone of the slip zone;on this basis,the formation mechanisms of two types of typical sandy loess landslides,transient and hysteretic,have been revealed by integrating the results of geological investigations,laboratory tests and numerical simulations.
Keywords/Search Tags:Sandy loess landslide, Landslide InSAR identification, Mechanical properties of slip zone soils, Landslide formation mechanism
PDF Full Text Request
Related items