【作者】 邢玉东；

【导师】 朱浮声；

【作者基本信息】 东北大学 ， 结构工程， 2008， 博士

【摘要】 论文以辽宁西部黄土为研究对象,以高速公路建设工程为依托,采用现场调查、测试和室内试验相结合的研究方法,通过对铁阜高速公路阜新一朝阳段湿陷性黄土的大量现场调查和勘察,利用三轴试验、14C测年法、X衍射法、扫描电子显微镜微观结构测试技术及计算机图像处理技术、人工神经网络技术、分形几何理论及基于扰动状态概念的建模方法,对辽西黄土的分布特征和工程特性进行了全面系统的研究和深入的探讨,开展了湿陷性黄土地基处理的试验研究,分析对比了试验前后黄土湿陷性的变化和黄土物理力学性质的改变,获得了各种湿陷性黄土地基处理方法的技术参数,为该地区工程项目的开展提供了科学的依据。论文的主要工作和成果如下：(1)阜朝高速公路沿线黄土分布于山前河流冲洪积平原、丘陵及低山地貌单元,主要分布于山前坡洪积扇及河谷阶地上。黄土为马兰黄土。黄土地貌形态特征并没有西北地区的黄土地貌特征明显,即没有典型的塬、梁、峁。不同地段的黄土均具有湿陷性,但不同地貌特征的黄土湿陷性具有一定的差异。在成因上可以认为是以风成为主后经重力和流水作用的综合成因,为冲洪积和坡洪积次生黄土。该地区黄土在颗粒成分上以粉粒为主,且其中以粗粉粒为主,粘粒含量变化较大,在粒度成分上有别于西北地区的黄土。黄土中的碎屑矿物以石英和长石为主,重矿物含量很小。粘土矿物以伊利石、伊蒙混层为主,绿泥石次之,高岭石含量很少。黄土中易溶盐平均含量为0.047%,中等溶盐平均含量为0.029%,难溶盐碳酸钙平均含量为2.46%,与我国其他地区的黄土中的可溶盐含量相比,易溶盐和难溶盐偏低。黄土的PH值在6.91-8.06之间变化,平均为7.47,呈弱碱性。(2)黄土的骨架颗粒的单粒状、集粒状和凝块状三种形式共存,粒状颗粒中单个的粉粒和由粘胶微细碎屑胶结成的集粒居多。黄土中粘粒含量较低时,骨架颗粒呈支架—镶嵌接触,随着粗颗粒含量的降低及粘粒含量的增高,骨架颗粒将由支架—镶嵌接触转变为分散分布,土体的强度和稳定性也随之增高。浅层黄土中大孔隙和中孔隙数量较多,说明存在较多的支架孔隙和其它大孔隙,而深层黄土的小孔隙和微孔隙则较多。黄土中孔隙具有自相似性,黄土中的孔隙分布具有分形特性,孔隙分布分维数在1.22-1.41变化。浅层黄土中的孔隙分布分维数较深部黄土略大。在深度20m内黄土的结构类型以支架大孔微胶结结构、支架大孔-镶嵌微孔半胶结结构、絮凝状胶结结构和凝块状胶结结构四种类型为主。浅部黄土的结构较松散,存在支架大孔微胶结结构,深度黄土的结构较为密实,为絮凝状胶结结构和凝块状胶结结构。由此,随着深度的增加,土体从松散逐渐趋于密实,土中孔隙逐渐减少,土颗粒从粒状逐渐聚集成凝块状,土粒间排列从微胶结结构逐渐变为胶结结构,稳定性逐渐提高,湿陷性逐渐降低。(3)黄土物理性质指标的变异性小于力学性质指标的变异性,力学指标离散性大。黄土的物理指标之间以及物理指标与力学指标之间相关性不好,相对而言,物理性质指标之间相关性要好于力学性质指标与物理性质指标之间的相关性。原状黄土饱和后的渗透系数大致为：竖向渗透系数的变化范围为4.13×10-5cm/s～9.13×10-5cm/s；水平向的渗透系数变化范围为1.67×10-5m/s-5.33×10-5cm/s；kz/kx值的变化范围1.54～4.65,渗透系数与渗透压力之间具有很好的幂函数关系,竖向渗透系数随深度具有较明显减小的规律。(4)辽西地区黄土具有有利于产生湿陷的基本原因,即特殊物质成分和特殊结构特征。深度1-5m的黄土的湿陷系数与物理指标之间的相关性并不明显。将人工神经网络方法用于黄土湿陷性判别,可以较好地利用物理指标进行湿陷性的预测。采用割线模量法建立了黄土湿陷变形的计算公式,可以用于湿陷性黄土地基的湿陷变形的计算。(5)在不同的试验条件试验获得强度指标是明显不同的,主要表现为不排水剪的抗剪强度指标c和(?)都明显较固结不排水剪获得的强度指标高。建议实际工程中采用有效应力强度指标。不排水剪的应力—应变关系主要表现为明显的应变软化型,但当固结压力较高时,应力—应变关系曲线变成应变硬化型；固结不排水剪获得的应力—应变关系主要呈弱应变软化型和应变弱硬化型；固结排水剪的应力—应变关系曲线为应变硬化型。这种强度和应力—应变特征主要受到黄土结构性、饱和程度、排水条件的影响。建立了描述黄土应变硬化和应变软化的扰动状态本构模型,可用于黄土路基的变形和稳定性分析。(6)冲击碾压法是处理浅层湿陷性黄土地基的一种有效方法,它可以增加地基土的干密度(密实度),增加地基土的承载能力,消除湿陷性黄土的湿陷性。冲击碾压的处治效果与湿陷性黄土的性质及设备冲击能有关,具体地段施工的效果应采用与施工地段地质条件相仿的路段试验确定。对于朝阳地区湿陷性黄土,25KJ三边形冲击碾压两试验段以12km/h冲碾速度碾压最优冲击遍数为40遍,冲击碾压的有效影响深度为0.90m。冲击碾压处治效果可采用动力触探的击数变化加以检验。强夯法作为一种处治黄土湿陷性的有效方法,适合于辽宁西部地区的湿陷性黄土的处治。在同一场地条件、试验设备情况下,单点夯击次数的不同,处治的有效影响深度不同,单点夯击次数大,影响深度越大,但变化不大。经强夯处理后的湿陷性黄土,消除湿陷性后土的干密度值大于1.50 g/cm3,可以采用土的干密度值作为检测的参考指标。湿陷性黄土强夯前后的动力触探击数变化与土的密实度变化有较好的关联性,可以用作判定处治有效深度的技术指标。通过采用灰土挤密桩可以有效消除黄土湿陷性,但消除的效果与采用的设备和设计的桩间距有关；可以用试验前后的动力触探对比判定场地土的处治效果；灰土桩处治施工后黄土的干密度值一般大于1.5g/cm3,可将干密度值作为黄土湿陷性检测判定的参考指标。更多还原

【Abstract】 Based on project of Fuxin-Chaoyang expressway, which is a section of Tieling to Chaoyang highway, the engineering characteristics of loess of west area in Liaoning is studied in this paper by the methods of field investigation, tests in-situ and tests in laboratory. So a lot of survey works and triaxial tests in laboratory have been done, then some advanced methods such as age test with 14C, X radial diffraction, test technique of scan electron microscope (SEM), computer image process technique, technique of artificial nerve network, theory of fraction geometry and modelling method based on disturbed state concept, are used to study and discuss systemically on loess distribution and its engineering properties. Furthermore, a great deal field tests are also studied on subgrade treatment of collapsible loess. Some technical parameters are obtained for collapsible loess subgrage treatment by contrasting the indexes after treatment and before treatment, such as collapsibility and physical and mechanical properties. All these results can provide scientific basis for future projects in this area.The main studies and creative results of this paper are shown in the following:(1) The loess in west area of Liaoning is mainly distributed in three unit of physiognomy, namely, the physiognomy of river alluvial, alluvial and flood plain and front area of mountain, the physiognomy of low mountain upland by tectonic erosion and the physiognomy of low mountain by tectonic erosion. The loess belongs to Malan loess. Loess geomorphology is characterized by no typical Yuan, Liang and Mao, which is significantly different from loess geomorphology of northwest in China. The loess in different location all have collapsibility, but there are some differences between different geomorphological units. From the view of origin, the loess is formed by winding and gravitational conveying and flow transporting, therefore the loess in this area can be regarded as secondary loess by alluvial-flood and slope-flood deposition.Grain composition of loess consists mainly of silt particle, in which coarse silt grain is great. The content of clay grain varies in more large extent. So there is different in grain composition from northwest loess. Clastic minerals are mainly quartz and feldspar, heave mineral is little. The main clay mineral is illite and mixture layer of illite and montmorillonite, secondary composition is chlorite, and kaolinite is little in the loess. The average quality content of easy-soluble salt in loess is 0.047%, medium-soluble salt is 0.029% and difficult-soluble salt (CaCO3) is 2.46%. The value of PH varies between 6.91 and 8.06, and the average of PH is 7.47, that is, loess is in a state of alkalescence.(2) Framework grain in loess is in a state of coexistence of three types of single, aggregate, and clot grain, and single grain silt and aggregate is more which cemented by colloidal and fine clastic particle. When content of clay mineral is low, framework grain will be in a type of bracket-inlay contact, and which will change into disperse type with decreasing of coarse grain and increasing of clay grain, so the shear strength and stability of loess will be improved. The number of large and medium pores of loess in a shallow depth is much more than in a deep depth, and the number of small and micro pores much more than shallow loess. It shows that shallow loess is looser, in which exists a lot of bracket pores and other type pores, on the other hand, deep loess is denser, in which exists more lots of micro and small pores. The pore of loess has a characteristics of self-comparability and its distribution has a characteristics of fraction geometry. Fraction dimension of pore distribution varies between 1.22 and 1.41, and this index in shallow loess is slightly greater than in deep loess. In depth of 20m, four main types of structure are (A) bracket and large pore with weak-cemented structure, (B) bracket large pore and inlay and micro pore with half-cemented structure, (C) flocculent-cemented structure and (D) clotted-cemented structure respectively. The structure of loess is looser in shallow depth, in which is in structure of A, and denser in deep layer, in which is in type of C and D. Therefore, it is shown that with the depth increasing, the structure tends to dense, the number of pores is getting to decrease, the type of grain is getting to change from single to aggregate, cemented characteristics among particles varies from weak-cemented into cemented type, so stability of soil enhances and collapsibility reduces gradually.(3) The variabilities for physical property indexes of loess are less than mechanical property indexes, that is, the great scatter distribution for the latter. Further analysis results express that there is neither a very good statistical relationship among physical property indexes, nor relation between physical and mechanical property indexes, and comparatively the former is slightly better. The coefficient of permeability for undisturbed loess sample after saturated is about from 4.13×10-5cm/s to 9.13×10-5cm/s in vertical direction and from 1.67×10-5 cm/s to 5.33×10-5 cm/s in horizontal direction. The ratio of kz to kx is between 1.54 and 4.65. There is a power relation between coefficient of permeability and seepage pressure. The vertical permeability coefficient decreases more obviously with depth.(4) The loess has basic reasons which lead to collapse, namely, special composition and special structure characteristics. It doesn’t have a very obvious statistical relation between coefficient of collapsibility and physical property indexes. It is found that technique of artificial nerve network (ANN) can used to predict reasonably the collapsibility of loess according to physical indexes. Here an expression which based on method of secant modulus also suggest to calculate rationally collapsible deformation of loess subgrade.(5) The indexes of shear strength determined from different test conditions are different. Results show that the c andφfrom the unconsolidated-undrained shear test (UU) are all larger than the indexes from consolidated-undrained shear test (CU). So indexes of effective shear strength is suggested to adopt in practical project. The stress-stain relation is mainly in a stain-soften pattern under UU condition, but will be transmitted to a strain-harden pattern when consolidated pressure is large. The stress-stain relation is mainly in a little stain-soften and little strain-harden pattern under CU condition, and it is a stain-soften relation under consolidated-drained test condition. These characteristics of stress-strain relation are influenced by structure, degree of saturation and drainage condition. Then a constitutive model, which can describe the characteristics of strain-harden and strain-soften, is established which can calculate deformation and stability of loess subgrade.(6) Impact-rolling method is an effective method to treat shallow collapsible loess which can increase dry density of soil and improve bearing capacity of subsoil, and eliminate the collapsibility of loess. The effect of impact-rolling method is related with property of collapsible loess and impactive effort, so the treatment effect should be determined by tests in a section of highway with similar geological condition. For Chaoyang collapsible loess the effective influence depth of impact-rolling can reach about 0.90m under impacting and rolling 40 times in the velocity of 12km/h with triangle impact-rolling machine in power of 25kJ.As a effective ground treatment method, dynamic compaction is propitious to treat the collapsible loess of west area in Liaoning. In the same field condition and test equipment, different compacted times, effective influence depth is different. The more of compacted times, the greater of influence depth, but in small variation. The value of dry density of soil can be greater than 1.50g/cm3 after collapsibility is eliminated. So the dry density can be reference index to examine the compacted effect. And there is a better relationship between number of dynamic penetration and degree of dense, so the number can be used as a technical index to determine the effective depth.Lime soil pile can effectively eliminate collapsibility of loess, but its effect relates with equipment and pile spacing. Treatment effect of this method can also be judged by comparing number of dynamic compaction before compact and after. The value of dry density of soil can be greater than 1.50g/cm3 after treating with lime soil pile, and can take dry density as a reference to examine collapsibility which is eliminated or not.更多还原