Study On Slope Stability Analysis Method Based On Reliability Theory

With the development of fundamental construction, there are more and moreobvious issues about slope stability involving civil, municipal works, water conservancy,highway, railway and mines. Especially in the terrain, geological conditions, large areasof undulating terrain, slope stability problem is particularly prominent. The traditionalmethod of calculating slope stability analysis is deterministic, but there are a lot ofuncertainties and unknown factors in the actual project, the safety factor method can notreflect the impact of the above variability. So, it is more conform to reality that evaluatethe slope stability by using the reliability theory.In this paper, reliability analysis of slope stability is presented using first-ordersecond-moment method(FOSM), and write two procedures based on the Swedish slicemethod and the simple Bishop method by MATLAB. The results of these procedures arecompared using Monte Carlo simulation method(MCSM), and compared the impact ofthe uncertainty of soil parameters and the distribution of soil parameters on thereliability of slope stability.This paper analyzes showed that:1A good agreement is observed between the calculated reliability index usingFOSM and MCSM for both the Swedish slice method and the simple Bishop method,the error does not exceed5%. So FOSM method used in the reliability of slope stabilityanalysis is feasible and effective.2When the FOSM based on the Swedish slice method, the reliability index is10%lower than the one based on the simple Bishop method. Compared to the results ofMonte Carlo method, it is showed that using the simplified Bishop method is morereasonable than the Swedish slice method.3The uncertainty of soil parameters has a great influence on reliability index. Withthe variation of parameters increases, the reliability index decreased rapidly, but thesafety factor method can not reflect this change.4The distribution form of soil parameter also has influence on reliability index. Inthe case of small sample and low uncertainties, the form of the distribution of theparameters could choice the student-T distribution. In the case of high uncertainties, theform of the distribution of the cohesion could choice the normal distribution or lognormdistribution and the form of the distribution of the angle of internal friction could choice the student-T distribution or normal distribution.