Development And Verification Of Rainfall Infiltration On Loess Landslides Stability Model

There are many types of landslides in the Loess Plateau.Rainfall is the main cause of loess landslides.Rainfall landslides have seriously affected the safety of human life and property.In order to avoid and reduce the casualties and economic losses caused by landslide disaster,the numerical model of landslide stability has become a hot topic.With the gradual deepening and improvement of numerical calculation theory and the continuous improvement of computer hardware conditions,numerical simulation has become an important method to understand and predict landslide instability.The numerical model of landslide stability considering rainfall can not only improve the prediction results of spatial probability of landslide occurrence,and then improve the prevention and control efficiency of regional landslide disaster in advance,but also strengthen the research on the instability mechanism of loess landslide,and provide some reference in the research fields of disaster prevention and reduction.Based on the existing research,firstly,based on the seepage equation and limit equilibrium method,this paper forms a specific calculation process from infiltration to stability,and establishes a complete model for rainfall landslide prediction.Then,using Kozeny Carman model and transformation function method,according to the physical and chemical parameters of loess,a method for predicting saturated permeability coefficient and soil water characteristic curve is proposed.Then,other software and classical examples are used to verify the developed infiltration and stability program respectively,and it is found that the model is feasible and correct.Finally,take zhongyanshan landslide as the object,simulate the two typical data,and finally assume the rainstorm condition to quantitatively study the impact of rainfall on the landslide.In the study of hydrological parameter modeling,other loess data are collected to verify the reliability of the parameter prediction model.At the same time,the relationship between liquid limit or average particle size and specific surface area is established,which proves that liquid limit and average particle size can also be used to predict saturated permeability coefficient.In the prediction,it is found that the difference of soil structure controls the hydrological parameters,and the soil structure is manifested in physical and chemical parameters,so the physical and chemical parameters are closely related to hydrological parameters.In the verification and application of loess landslide stability model under the influence of rainfall,firstly,the feasibility and correctness of this model are verified by typical examples and other software.Then,based on the stratum data and meteorological data,the volume water content distribution in the soil in the two marked periods is calculated according to the rainfall and solar radiation,and compared with the actual data,it is found that the calculated results are consistent with the actual results.Then,according to the pore water pressure distribution in each period,calculate the changes of the most dangerous sliding surface and landslide stability,so as to achieve the purpose of predicting landslide stability.Finally,the unstable rainfall threshold of zhongyanshan landslide is inferred by setting the rainstorm condition and calculating the variation law of safety factor with time.The calculated rainfall threshold is 600 mm.In the study on the change of the most dangerous sliding surface,it is found that rainfall is easy to induce larger and wider landslides.For landslides with small thickness,the instability is mainly due to the increase of pore water pressure of surface soil.For landslides with large thickness,the instability is mainly due to the rise of groundwater level.In this paper,the model is established based on seepage equation and limit equilibrium method,and the hydrological parameters are modeled.After verifying the effectiveness of the method,the zhongyanshan landslide is taken as the research object for numerical simulation.This study is of great significance to the understanding and prediction of rainfall loess landslide.