【作者】 Hussein Yousif Aziz；

【导师】 Jianlin Ma；

【作者基本信息】 西南交通大学 ， 桥梁和隧道工程， 2013， 博士

【摘要】 高速铁路对地基沉降控制提出了严苛的要求,同时桥梁深基础的沉降计算有很多不准确的因素。因此,有必要结合现场试验对计算方法进行研究。本文采用了京沪高速铁路某桥梁试验点的单点沉降仪和液位沉降计的监测数据。文中介绍了一种获取压缩层厚度的新的测量方法,在该方法中巧妙的克服了单点沉降计无法稳定地锚固在岩土体中进行有效监测的困难。试验结果表明,压缩层厚度取决于土的特性。本文采用神经网络系统(NFS)、双曲线模型和统计回归模型对桥梁桩基沉降随荷载的变化进行预测,并与实测值进行对比研究。结果表明,神经网络系统(NFS)对桩基沉降的预测结果优于双曲线模型和统计回归模型,该方法可以应用于岩土工程中类似的或更复杂的问题中。本文提出了一个压缩模量的修改公式,可以对土工试验压力为100～200kPa下深厚软土的压缩模量进行修正。结合相关规范和模量修正公式,对桩基沉降进行计算,结果表明,模量修正公式可以提供计算结果的准确性。采用应力控制法来控制压缩层厚度,计算深长摩擦群桩在深厚软土中的沉降。与实测值相比,应力控制法结合修正后的土体模量的计算结果与实测值较为接近,表明该方法适用于深厚软土地区的沉降计算。结合现场试验工点的地质情况,本文对比分析了三种国内外桥梁桩基规范：美国规范(AASHTO code),建筑地基基础设计规范(GB50007-2011)和铁路桥涵地基和基础设计规范(TB10002.5-2005),以找到运营荷载作用下,桥梁桩基沉降计算最优的规范。并且对单桩的竖向极限承载力进行了研究。结果表明,AASHTO规范对单桩承载能力的计算更加安全,而建筑地基基础设计规范(GB50007-2011)对地基沉降的计算结果更为合理。采用修正后的土体压缩模量,应用Plaxis3D Foundation进行有限元分析,结果表明修正模量结合Plaxis3D是进行数值模拟的一个有效手段。采用修正后的土体压缩模量的数值分析的计算结果更接近实测值。相比双曲线模型和统计回归模型,有限元数值分析结果更为准确。本文最后对位于伊拉克南部穆萨纳省的Al-Rumaitha桥进行了研究。结合实际地质情况,分别采用规范和Plaxis3D和2D有限元进行沉降计算,三者计算结果基本一致,并且可用模量修正公式对沉降进行修正。承载能力的计算结果表明,桥梁的承载能力大于卡车通行情况下的运营荷载,说明该桥可用作高速公路。更多还原

【Abstract】 The design of high speed railways need for very strict requirements of settlement control. The settlement calculations of deep bridge foundation suffer from many obstacles and inaccuracy. Therefore, there is a need for the accurate methods of calculations to match the reality in the field.The study of measuring the settlement values of the Bridge in Beijing-Shanghai High-speed Railway with the time and load is considered in this dissertation. The settlement is measured using the single-point settlement gauge and hydrostatic level. The results of field test show a new field measurement method to calculate the compressed layer thickness. A single settlement account with a single point measurement is used to overcome the difficulties, when the base bottom layer cannot be stable anchorage in the soft soil or rock layer be difficult to monitor. Field tests showed that the deep pile support layer compression thickness is closely related to soil properties.This dissertation is also a study on the behavior and prediction of settlement values of bridge pile foundations due to construction loads. The settlement is predicted using hyperbolic model and statistical regression. The field measurements are compared with the hyperbolic model results and those predicted by the statistical regression. The Neurofuzzy System (NFS) used in this dissertation to predict the settlement of deep pile foundations. The results obtained from this system give good agreement and high accuracy for prediction of settlement compared with hyperbolic model and statistical regression analysis. Also, this scenario can be applied for similar or more complicated problems in the Geotechnical engineering.A new formula for compression modulus modification that calculated from compression modulus provided by geotechnical test under pressure of100-200kPa along thick soft soil is proposed in this dissertation. Based on the new formula, comparative analysis of group piles settlement by existent national specifications is carried out. This finding modifies the codes calculations which are lack the accuracy and confidence of settlement calculations in the deep soft soils. Method of stress control is used to control compressive layer thickness, and adopted to calculate settlement of long friction group piles in thick soft soil. Through comparison with the field data, settlement values by this method using the modified compression modulus are close to the measured values, which prove the applicability in calculating settlement in thick soft soil.The dissertation also discusses the design and analysis of a bridge foundation subjected to load of train with three codes, namely AASHTO code, Code for Design of Building Foundations (GB50007-2011) and Code for Design on Subsoil and Foundation of Railway Bridge and Culvert (TB10002.5-2005). The study focuses on the design and analysis of bridge foundation and find which code is better to control the problem settlement due to the applied loads. Settlement of the bridge is calculated depending on the data collected from the project site. The vertical ultimate bearing capacity of single pile for the three codes is also discussed. The results indicate that the AASHTO code is safer in the bearing capacity of single pile, while Code for Design of Building Foundations (GB50007-2011) gives a good indicator of the risk to foundation settlement.The finite element Plaxis3D Foundation program is used in the analysis with the new empirical equation to modify the input parameters represented by the soil compression modulus. The results of the numerical analysis using the proposed empirical equation provide insight to the settlement analysis of pile groups in soft clayey soils and the finite element Plaxis3D program can be a useful tool for numerical analysis. In this dissertation, the numerical analysis calculations are modified using the correction formula to calculate the compression modulus from those obtained in the test which modify the results of the settlement and thus become close to the reality. The numerical finite element analysis produced accurate results as compared to the statistically derived equations and those calculated by hyperbolic model analysis.This dissertation finally studies on the Al-Rumaitha Bridge lies in the Muthanna province, south of Iraq, including the soil investigation taken from the project and settlement analysis for the designed bridge firstly manually according to the data provided, secondly using two and three dimensional Plaxis Foundation program of finite elements. The settlement calculated by the above three ways is nearly similar, and the results can be modified using the proposed formula in this dissertation. The bearing capacity of the bridge is also calculated and its value is larger than the applied load due to the effect of truck loading. The study found that this bridge can be used in the highway and can success for this purpose.更多还原