Study On Stability And Antiknock Characteristic Of High Fill Subgrade Sides In G577,Xinjiang

In recent years,some accidents such as collapse and landslide always occurred in high fill embankments,which show great relationship with unreasonable embankment design methods.At the present stage,limit equilibrium method and slices method show two obvious disadvantages though they had advantage of simple formulas.(1)Limitation in evolution of failure probability for high fill embankments;(2)Unreasonable calculation results of earthquake force,especially horizontal earthquake force.In order to solve the two problem,side failure probability model and improved quasi-static method were put forward combined with the national road G577 from Hantian to Tekesi.A method for calculating safety factor was presented regarding antiskating force and sliding force as two independent random variables.Based on this,a calculation method for reliability was put forward to compute the probability rate of damage in high fill subgrade slide and utilized in the national road G577 from Hantian to Tekesi.Meanwhile,the main disadvantages of traditional quasi-static method were discussed.Based on the discussion,an improved quasi-static method was put forward by using horizontal slices method,which overcame the main disadvantages of traditional quasi-static method and provided a theoretical basis on engineering.At last,combing with landform,climate,water distribution,hydrogeological conditions,geological structure and lithological characters in the national road G577 from Hantian to Tekesi,the calculation method for reliability and the improved quasi-static method were utilized to verify their accuracy.The results showed that side failure probability model could evoluate failure probability.Meanwhile,only in the conditions whose safety factor was more than 1.2 and failure probability was less than 1.25%,high fill embankments wouldn't break down in their life cycles.improved quasi-static method could calculate safety factor more accurate by avoiding using unreasonable value of ai? In future,more practice should be done to evaluate the accuracy of these two methods.