基于局部随机场有限元理论的古丈县二龙庵滑坡分析

发布时间：2018-06-22 23:03

本文选题：二龙庵滑坡 + 空间变异性　；参考：《北京交通大学》2016年硕士论文

【摘要】：边坡稳定性一直是岩土工程研究领域一个重要问题,由于岩土体材料具有明显的空间变异性,即使是单一的土层,不同点之间的土性参数特性也存在着一定的差异,以至于基于确定性分析的边坡稳定性计算结果价值有限。针对这个问题,考虑了岩土体土性参数的空间变异性,开发了基于蒙特卡罗(Monte-Carlo)法的随机有限元分析程序,对湖南古丈县二龙庵滑坡的安全系数和失效概率以及滑坡诱因进行了深入的分析,并对土性参数空间变异性影响条件下边坡稳定性规律进行了探讨,具体研究内容和成果如下：(1)调研了国内外关于岩土体空间变异性、随机有限元分析以及边坡可靠度分析的相关研究,并对这些研究进行了分类。(2)基于极限平衡法和蒙特卡罗(Monte-Carlo)法进行了边坡的可靠度分析,研究了土体参数的变换变异水平和分布类型对滑坡的影响。可靠度分析结果表明：大致以边坡失效概率Pf=50%为界限值,当Pf50%即界限值以上随着变异系数不断增大失效概率不断减小,当Pf50%即界限值以下随着变异系数不断增大失效概率不断增大；同时要注意方差过大时导致正态分布类型下随机数生成出现的负值以及对数正态分布类型下概率密度函数图像偏移导致生成整体值偏低的情况。(3)由于自然条件下,岩土介质具有明显的空间变异性,结合空间变异性的相关指标(方差、方差折减函数、相关函数、波动范围等)对均质边坡进行了随机有限元分析。结果表明：对于不同的相关距离和方差,土性参数的空间分布和边坡滑裂面对比具有很大的差异性,表明边坡稳定性分析时考虑空间变异性具有重要的实际意义。(4)基于蒙特卡罗(Monte-Carlo)随机有限元法和有限元强度折减技术对湖南古丈县二龙庵滑坡进行了局部随机场展开和可靠性分析。结果表明：与粘聚力相比,滑裂面所在土层的内摩擦角对滑坡的影响更为显著,其中第2层土对滑坡影响最大,并且数值模拟出的潜在滑裂带与实际勘察滑裂带结果吻合度较高。表明基于随机场展开的有限元分析和有限元强度折减技术在边坡稳定性分析中具有重要的应用价值。
[Abstract]:Slope stability has always been an important problem in geotechnical engineering. Because of the obvious spatial variability of rock and soil materials, even in a single soil layer, there are some differences in the characteristics of soil property parameters between different points. Therefore, the calculation results of slope stability based on deterministic analysis are limited. Considering the spatial variability of soil and rock parameters, a stochastic finite element analysis program based on Monte-Carlo method is developed. The safety factor, failure probability and landslide inducement of Erlongan landslide in Guzhang County, Hunan Province are deeply analyzed, and the slope stability law under the influence of spatial variability of soil parameters is discussed. The specific research contents and results are as follows: (1) the research on spatial variability of rock and soil mass, stochastic finite element analysis and slope reliability analysis has been investigated at home and abroad. These studies are classified. (2) based on the limit equilibrium method and Monte-Carlo method, the reliability of slope is analyzed, and the influence of variation level and distribution type of soil parameters on landslide is studied. The results of reliability analysis show that the limit value of slope failure probability is about 50%, and when the limit value is above 50%, the failure probability decreases with the increasing of coefficient of variation. When Pf50% is below the limit value, the failure probability increases with the coefficient of variation. At the same time, we should pay attention to the negative value of random number generation under normal distribution type when the variance is too large and the probability density function image migration under logarithmic normal distribution type, which results in low overall value. (3) under natural conditions, The spatial variability of rock and soil media is obvious. The stochastic finite element analysis of homogeneous slope is carried out by combining the related indexes of spatial variability (variance reduction function correlation function range of fluctuation etc.). The results show that for different correlation distance and variance, the spatial distribution of soil parameters and slope slip face ratio have great differences. It shows that it is of great practical significance to consider spatial variability in slope stability analysis. (4) based on Monte-Carlo stochastic finite element method and finite element strength reduction technique, the local random field development of Erlongan landslide in Gulang County, Hunan Province is carried out. Open and reliability analysis. The results show that the influence of the angle of internal friction on the landslide is more significant than that of the cohesive force, and the second layer of soil has the greatest effect on the landslide. The potential slip zone obtained by numerical simulation is in good agreement with the result of actual investigation. It is shown that the finite element analysis based on random field expansion and the finite element strength reduction technique have important application value in slope stability analysis.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P642.22

【相似文献】