【作者】 孙军杰；

【导师】 王兰民；

【作者基本信息】 兰州大学 ， 地质工程， 2010， 博士

【摘要】 震陷是地基基础的主要震害形式,黄土震陷是黄土地震灾害的重要类型。目前,对黄土震陷的时域发展过程了解甚少,黄土震陷与土体结构强度之间的定量关系研究尚属空白,黄土场地震陷引发的桩基负摩阻力的基本特征仍待研究。黄土震陷相关研究的难点在于实际震例太少,现场试验是解决这一缺陷的有效手段,然而与此有关的现场试验研究工作或成果,鲜有文献报道。论文以微差爆破激发的场地强震动为基础,实施了针对黄土场地震陷与桩基负摩阻力的现场试验研究。结合现场试验和与之配套的室内动三轴试验等数据,开展了爆破模拟地震动作用下黄土场地地面运动特征、黄土震陷的场地分布特征与时域发展规律、黄土震陷数学统计模型与概率性预测方法和黄土震陷时桩基负摩阻力衍生特征与生成机理等诸项研究工作,取得了如下创新性成果。(1)通过爆破地震动模拟效果分析,发现爆破模拟地震动的有效持时是影响震源设计效果的主要因素,爆破产生的强地面运动具有持时任意叠加的特性,据此给出了基于持时与有效持时的爆破地震动合成时程,对该时程进行简单接续能够获得用于满足不同研究需求的输入动荷载。(2)综合对比分析了爆破模拟地震动和黄土震陷之间的场地特征分布,指出地震动强度与频谱H/V值的耦合分布特征,对黄土震陷的场地分布形态具有显著影响,场地内耦合量值较高的区域具有较大的震陷量,反之亦然；同时,地形地貌条件也是控制黄土震陷的场地分布特征的重要因素,沉降分布的对称轴随观测时间延续逐渐向场地土体约束较小的南侧陡崖处偏转。(3)沉降观测结果显示,爆破结束时场地土体震陷量可达沉降终值的50%,其后将会有一个相对缓慢的增长过程；通过统计分析爆破模拟地震动作用下黄土震陷的时域发展规律,给出了黄土震陷量随爆破地震动能量输入快速累积和后期土体因结构强度损伤再固结持续发展的联合时域过程。(4)结合试验研究结果与物理力学机制分析,将黄土场地震陷的主导因素归纳为表征土体强度特征的粘聚力(C)、内摩擦角(φ)、表征土体沉降特征的孔隙比(e)、表征外部输入动荷载特征的地震动强度(PGA)与频谱特性(如H/V值)以及场地地形地貌条件等；场地条件与土体物性(或输入动荷载)对黄土震陷的影响机制不同,在具体分析中可独立地进行考虑。(5)由黄土震陷物理力学机制入手,给出了利用孔隙比描述震陷系数的物理定义,发现了孔隙比压缩量和以动应力与固结应力为基础定义的应力比之间具有较好的统计关系,建立了基于黄土震陷物理力学机制的数学统计模型；在此基础上,得到了黄土震陷系数概率预测模型的三维理论矩阵,结合计算得来的黄土高原地区地震动重复概率,提出了区域黄土场地震陷的概率性预测方法。(6)分析研究了爆破地震动作用下的桩体形变响应特征,认为桩身轴力最大处是形变的最大响应部位,这一位置与桩基负摩阻力中性点深度相对应；现场试验期间,中性点深度随负摩阻力发展时间的推移略有加深,但整体变化幅度很小,可近作桩身中性点的位置基本固定。(7)现场试验监测结果表明,黄土场地震陷引发的桩基负摩阻力沿深度方向逐渐增大,在邻近中性点处(负摩阻力为0kPa)达到最大量值；桩身负摩阻力的平均值可达54kPa,对应的总负摩阻力约为1654kN,远高于以往浸水湿陷条件下的测试结果；综合负摩阻力具有的较高量值、沿桩身分布形态和中性点深度基本不变等特征,得出了该负摩阻力与静摩擦力学特性相关联的结论。(8)根据桩基负摩阻力衍生特征、桩身偏心应力(桩体两侧摩阻力差值)状态和已有负摩阻力测试资料,系统分析了桩基负摩阻力的生成机理,指出桩基负摩阻力具有沉降相关性、强度相关性和静动摩擦力学相关性等3大特征属性；场地沉降发生期间,场地土体结构强度是影响桩基负摩阻力量值的决定性因素。更多还原

【Abstract】 Seismic subsidence is a principal kind of ground disaster due to seismic loading, and seismic subsidence of loess (SSL) is the important one of damage behaviors for this special type of soil caused by earthquake. As we known, essential developing features in time domain of SSL, quantitative relations between SSL and structure strength of the soil mass, as well as primary characteristics of negative skin friction (NSF) along piles induced by SSL, are still indistinct. The crucial challenge, unfortunately, we are facing on the research works associated with SSL, lies in the relevant cases are so infrequent that NSF caused by SSL is difficult to investigate and the study of SSL remains hard to go on with in the field of earthquake engineering. Field tests are the effective methods to eliminate the defects, whereas the existing in-situ data is rare to meet the need of the study.Based on the field test by a short delay blasting, this dissertation takes major efforts on essential characteristics of SSL and NSF associated with the soil settlement. Using the data of in-situ exploding experiment and laboratory dynamic triaxial test, the author investigates behaviors of loess ground induced by the short delay blasting, distributional and developing characteristics in time domain of SSL, mathematical statistical model and probability-based prediction method for SSL, and essential features and generation mechanism of NSF caused by SSL. The creative achievements under the above-mentioned research works are summarized as the follows.(1) The analysis results of the exploding ground motions reveal that duration are the most important factor to influence design effects of a short delay blasting. Independent exploding events in the blasting process could be linked one to another and then the duration of exploding ground motion may be lengthen. These results would not influence the spectral characteristics of ground motion induced by the short delay blasting. Those actual time histories of ground shock with longer duration are better in meeting the demands for laboratory and field tests to understand response behaviors of soils and structures.(2) Grounding on the comprehensive analysis in distributional characteristics of exploding ground motion and SSL, the author figures out the coupling characteristics of PGA and H/V values could make an appreciable impact on the distribution of SSL in the loess field. The area with both higher PGA and H/V values of spectral features for exploding ground motion could suffer a greater SSL, and vice versa. Meanwhile, geographic and geomorphic conditions could strongly affect distributional features of soil seismic subsidence in a loess ground. In this experiment, the symmetry axis of SSL’s distribution gradually rotates with time and directs towards the cliff that stands to the south of the field, where the soil bears a lesser settlement restriction, at the end of the in-situ observation.(3) By analyzing the developing regulation in time domain of SSL caused by the exploding ground motion, we can get that just at the end of the blasting, SSL could reach about 50% of the total observational settlement, while the other half would settle during a longer period later, i.e., SSL should experience a relative long-term-process during the field test, including two stages following one after the other, the rapid one and the tardily one. The rapid stage occurs along with the exploding event, whereas the tardily one is generated by the soil reconsolidation after structure strength is decreased due to the short delay blasting.(4) From the contrastive analysis results of field experiment, laboratory dynamic triaxial test and physical and mechanical mechanism of SSL, the dominant factors affecting SSL could be came down to as follows, cohesive strength (C), internal friction angle (φ), void ratio (e), ground motion intensity (PGA) and spectral features (e.g. H/V values) of ground motion, and geographic and geomorphic conditions. The factors of C and cp briefly represent the characteristics of soil structure strength, while e influences the final seismic subsidence magnitude of soil mass. Geographic and geomorphic conditions in the field are independent of soil’s physical parameters or of ground shock on the influence mechanism of SSL. The geographic and geomorphic conditions could be analyzed separately during the analysis of dominant factors influencing SSL.(5) Based on the physical-mechanical mechanism of SSL, the relation of void ratio and seismic subsidence coefficient (SSC) is established through physical and mathematical analyses. It is clear that a stress ratio (defined by dynamic stress and consolidation stress) coincides with the compression values of void ratio. Consequently, both mathematical statistical model of SSL and three-dimensional theoretical matrix for regional assessment of SSL are educed. Moreover, curves of seismic probability within different reoccurrence periods of years for PGAs in the Loess Plateau are calculated and the probability-based prediction method for SSL in the region of Loess Plateau is provided. (6) Strain behaviors of pile body in response to exploding waves indicate that the maximum response magnitude and axial force along pile appearance at the same position, which is the buried depth of neutral point also, where NSF is OkPa. During the field experiment, the depth of neutral point increases gradually after the short delay blasting, however, the increment is very small, i.e. the neutral point could be considered as a fixed position with no change.(7) The field testing data show NSF rises gradually with increasing depth and the mounts up to the maximum value near the neutral point. Furthermore, the final average NSF reaches 54kPa approximately, with a corresponding total NSF along pile of 1654kN. These results are much greater than the previous NSF cases induced by loess settlement due to water soaking. The special characteristics of NSF, higher values, distributional features along pile and almost fixed position of neutral point, reveal that NSF observed in this field experiment might be associated with static friction.(8) By the comprehensive analysis of NSF’s generating-and-developing features, pile body’s eccentric stress (differences of NSF between two sides of each pile) and existing NSF data, it could be affirmed that NSF shares correlations with 3 aspects as follows, ground settlement, structure strength of soil mass, and static or kinetic friction. During the ground settlement, generally, structure strength of soil mass is the foremost factor that influences NSF.更多还原