【作者】 高德彬；

【导师】 夏永旭； 胡长顺；

【作者基本信息】 长安大学 ， 道路与铁道工程， 2008， 博士

【摘要】 随着国家西部大开发的实施,作为带动西部经济发展的交通运输工程(特别是高速公路)已得到优先发展,尤其西部黄土地区公路建设更是得到迅猛发展。在地质背景复杂的黄土地区进行公路建设中,形成了大量黄土路堑高边坡。这些黄土路堑高边坡由于纵断面长、横断面宽、坡高极高,工程量巨大,在经济合理的前提下保证边坡的整体稳定性与坡面稳定性,是工程中需要解决的技术难题。同时,经过近十年高等级公路的运营,对大量黄土路堑边坡的使用状况已有初步的认识,这为研究黄土路堑高边坡的稳定性提供了借鉴。本文依托西部交通建设科技项目“黄土地区公路高边坡防护技术研究(200131800020)”,通过对黄土地区已建或在建公路黄土路堑高边坡的大量野外调查,结合室内外试验和理论计算分析等研究,取得如下主要成果:(1)依据黄土的成因时代、物质组成、结构构造及水文地质等条件,总结出了的黄土路堑高边坡的八大类地质结构类型模型。同时分析了黄土路堑高边坡坡面与坡体变形破坏的特征。这对公路黄土路堑高边坡稳定性分析以及防护工程设计都具有非常重要的实际意义。(2)将现场调查55个典型边坡作为范例,首次建立了以范例推理方法为手段的工程类比数据库,为拟建的黄土路堑高边坡设计提供了设计参考依据。该数据库将变权思想应用到属性权值的计算中,体现了权重对环境的敏感性,计算结果更加接近实际。(3)根据大量调查资料和现场冲刷试验结果,并结合理论计算,总结出了公路黄土路堑高边坡的适宜坡型(阶梯型)和坡面防护措施。同时,结合前人研究成果,提出了公路不同分区、不同地质结构模型的黄土路堑高边坡坡型设计和坡面防护方案推荐表。(4)通过实验发现,钻孔取样对土样有一定的扰动性,参数存在较大误差,本文根据探井取样结果研究,给出了黄土参数的修正公式。另外,根据大量参数统计分析发现:水对直剪试验强度指标的c值影响较大,对?值影响较小。同时经统计发现,黄土的物理指标重度的变异性小。抗剪强度指标c和?值的大范围统计变异性很大,单个工点统计变异性要比大范围统计的变异性小。当采用不同深度的强度指标进行统计分析时,可不考虑参数相关性的影响。物理指标含水量w和重度γ基本符合正态分布和对数正态分布,少数符合极值Ⅰ型分布。抗剪强度指标c和?大部分符合正态、对数正态和极值Ⅰ型分布,少数符合威布尔分布。(5)采用Fredlund非饱和土强度公式,对取自西禹高速某边坡的Q2黄土样,进行基质吸力测试,获得强度公式中相应的指标。并通过对含水量与基质吸力进行回归,得到二者的关系式,可依据含水量计算基质吸力,免去测试吸力的麻烦。(6)根据现场调查的251个黄土路堑边坡,从中选择207个典型稳定边坡,根据公路工程地质法的四个分区,八类地质结构模型,建立了坡高与综合坡度的对数线性相关关系。(7)建立了基于简化Bishop法的Monte-Carlo法的边坡可靠度指标计算方法。对四区八类地质模型共142个计算坡高点,12种强度参数变异系数组合,进行1704次的模拟计算。根据对失效概率Pf与可靠度指标β的变化分析,且考虑目前工程中测试黄土强度变异性的水平,取δc=0.3,δ?=0.15组合作为高等级公路黄土路堑高边坡失效概率推荐的控制因素,取δc=0.5,δ?=0.2作为二级及二级以下公路的失效概率推荐的控制因素,给出了不同分区、不同地质结构模型和不同等级公路的黄土路堑高边坡失效概率及可靠度设计标准推荐方案。更多还原

【Abstract】 With the development of the Western region, the communication and transportation has been given first priority to development, especially the highway construction in the loess area. A plenty of high cutting loess slopes were built in such a complicated region of the geological background. Because of the long vertical section,the width of the cross section,the vast high slope and the amount of work in the high cutting loess slopes, the stability of slope entirety and slope surface being guaranteed is a technical problem which is difficult to solve in the project under economical reasonable condition. And the behaviors of loess cutting slopes in service been preliminarily understood, the massive examples of high slope engineering have provided reference on studying high loess slopes when highways have been used for nearly ten years. With the backing of West transportation construction science and technology project named“The research on protection techniques of road high slopes in loess area”(200131800020),and by means of a vast amount of investigation data of high cutting loess slopes built or being built,and some experiment inside and outside,theory calculation. etc,the research results have been obtained some as follows in this paper:(1). Based on the origin age, component,structures and water geological condition,etc,8 kinds of geological structure type models were summed up in the cutting loess high slopes. Simultaneously,the feature of shape destroy was examined between surface and slope. these research has the extremely vital practical significance on slope stability analysis as well as the protection engineering design of high loess cutting slopes.(2). 55 typical slopes,which were taken for the case model, were firstly set up the project analogy database by means of Case-based Reasoning method in order to provide the most reference foundation on later slope design. The database, which made use of variable weight thought to calculate in the weight of attribute, gave expression to sensitive that exert weight influent on environment. The results of calculation were close to practice.(3). According to investigation of plenty of materials and the results of rainfall washout experiment on the spot, the suitable style in cutting loess high slope and the guard ways of surface were summed up in light of theory calculation. Slope protection plan recommendable table and the cutting loess high slope design were put forward by different zones and geological structure model with former results.(4). After testing, samples were be turbulent and exist a quiet error in parameter by drill holes technique. A revised formula was set up on the basis of man-excavated well samples testing. A plenty of parameter statistics suggested that water put more influence on the strength indexes c, but more than ? in direction shearing testing. And the variability of physical indexγis small. The large statistic variability of shearing indexes is enough big, but a project statistic variability of shearing indexes is less than it. When kinds of degree of depth in strength indexes were analyst by statistics, the influence of the parameter mutual relation was not considered. The physic indexesωandγsubmitted normal and logarithm normal distribution on the whole, only a few submitted extreme value I distribution. Most of the shearing strength c and ? submitted normal, logarithm normal and extreme value I distribution,a tiny minority submitted Weibull distribution.(5). Using Fredlund non-saturated soil strength formal and matrix suction test,the correlation indexes of strength formal from Xi Yu highway samples were obtained. After regressing between water content and matrix suction,their relation was determined. It was able to calculate matrix suction only using water content without the trouble of suction test.(6). After selected 207 typical stable slopes from 251 loess slopes on investigation and divided into 4 zones and 8 type geology structure models by railway engineering geology,the logarithm liner relation was defined between slope high and the comprehensive slope rate .(7). Based on the simplified Bishop method and Monte-Carlo method,the reliability index of the slope calculation method was established. Altogether 142 computation slopes high spot from 4 zones and geology models of 8 types were simulated calculation with 12 kinds of coefficient variation combination after 1704 times. After analyzing the changing between the failure probability and the reliability index , and given the level of loess strength variability in practical project,δc=0.3 andδ?=0.15 combination was taken as the controlling factor of first-class highway failure probability recommendation,δc=0.5 andδ?=0.2 as the controlling factor of two levels and two level of below roads failure probability recommendation. Finally,different zones, geology structure models and rank highway of failure probability design standard recommendable plan were given in loess cutting high slope.更多还原