重塑黄土直剪试验误差分析及对边坡可靠性影响评价

Error analysis of remolded loess by direct shear test andevaluation of influence on slope reliability

为通过室内快剪试验误差来源分析强度参数的随机特性,比较了击实法和压样法对抗剪强度离散性的影响,并在两种制样方式的基础上,进行再现性试验,通过对抗剪强度及正应力修正,采用Grubbs准则检验可疑值,求得了能反映强度参数随机特性的概率分布,并使用K-L级数展开法离散强度参数正态随机场,分析了两种制样方式对边坡可靠性的影响.研究表明:压样法制样比击实法获得数据离散性更小,使用再现性法能够体现强度参数的随机特性;两种制样方法在边坡稳定性分析时安全系数接近,由于击实法统计数据的离散性更大,可靠性分析时比压样法产生更大的失效概率.

In order to analyze the random characteristics of strength parameters through the error sources of indoor fast shear test,the influence of compaction method and pressure sample method on the dispersion of shear strength was compared,and on the basis of the two sample preparation methods,the reproducibility test was carried out.Through the correction of shear strength and normal stress,the Grubbs criterion was used to test the suspicious values,and the probability distribution reflecting the random characteristics of strength parameters was obtained.The influence of two sample preparation methods on slope reliability is analyzed by using K-L series expansion method to discrete the normal random field of strength parameters.The results show that the discreteness of the data obtained by pressure sample method is smaller than that obtained by compaction method,and the reproducibility method can reflect the random characteristics of the strength parameters.The safety factors of the two sample preparation methods are close to each other in slope stability analysis.Due to the greater discreteness of statistical data of the compaction method,the failure probability of the pressure sample method is greater than that of the compaction method in reliability analysis.