| The loess landslide is a common geological hazard, which often causes enormous loss to lives and property of the people, increases the risk of degradation of human environment. The sliding belt of loess landslides plays an important role to the loess landslides, the shear strength of sliding belt soil impacts the stability straightly, and the determination of it is the key to stability evaluation of a landslide. On the base of investigation of loess landslides, the shear strength parameters statistical rules of sliding belt soil are applied to the reliability analysis of loess landslides, the research work has been carried out, which including these three aspects: the geological background and influential factors of loess landslides, the statistics of shear strength parameters of sliding belt soil, the stability reliability evaluation of loess landslides.(1) The spatial distribution of loess landslides develop unevenly or highly concentrate in engerineering site.The tempo—law varies regularly with the corresponding characteristics of season, climate and engineering geology of the loess landslides region. The development of loess landslides is impacted by the factors such as formation lithology, topographty and physiognomy, geological structure, ground water and engineering activities of humankind.(2)The characteristics of sliding belt is analyzed, 32 sets shear strength tests parameters (cohesion c and friction angleφ) of sliding belt soil in 9 regions are summed respectively according to three kinds of sliding belt soil (pure loess, the mixture of loess and efflorescence of bedrock, weak formation of bedrock) and four different shear tests such as direct quickly shear, rest shear, saturated clotty shear and clotty shear. The bar graph and rules of shear strength parametric statistics are obtained which show the average value, standard deviation and coefficient of variation of c andφpossess the regional characteristics, and the distribution of c andφfit several different types, especially to normal distribution, log—normal distribution and the max—value I distribution.(3)Based on the theory of reliability, the reliability analysis model of loess landslide built on transmittance method is founded and the rules of parametric statistics of sliding belt soil are applied to the model. Furthermore, a typical case study was selected to verify the model, the failure probability and the distribution of stability factor are calculated by the way of Monte—Carlo; the effect of frequency in sampling , distribution and parameters estimation of average value, coefficient on failure probability are studied, which indicate that the more frequency in sampling is, the more accurate the result is, the result is safer while c andφfit the normal but more dangerous when fit the max—value I distribution, the failure probability increases with the rise of coefficient of parameters estimation and reduces with the rise of average values, the stability factor and failure probability of a loess landslide are more sensitive to the distribution, estimation of coefficient of variation and average value of friction angle 9.
|
PDF Full Text Request