Loess slope stability analysis based on reliability

This article takes the Hutoumao loess slope in Yan’an City as the research objective. The soil density r, cohesion c and frication angle φ of the soils of the slip zone were analyzed and determined as the random variables of the slope reliability on the basis of detailed field survey and exploration. Using the moment estimation method, first-order second-moment method, geometry method and Monte-Carlo simulation method, the slope reliability index and destruction probability were calculated separately. The safety factor is 1.00-1.15, reliability index is 0.35-0.45 and destruction probability is 32-37%. The results show that the slope is more likely to be unstable. The calculation process indicates that: the geometry method has the best reliability index precision, the first-order second-moment method is less precise, and the moment estimation method is least precise. The Monte-Carlo simulation method can check the results of others methods as a verifying method. The geometry method can be used when the fixed safety factor is not necessary, because it is relatively easy. But when there is a need to know the safety factor, the moment estimation method and the first-order second-moment method are in point.