节点文献
大跨度斜拉桥全寿命健康监测几个关键问题研究
Several Critical Problems Reasearch on Life Cycle Healthy Monitoring Method for Loangspan Cable- Stayed Bridge
【作者】 杨小森;
【作者基本信息】 北京工业大学 , 土木工程, 2011, 博士
【摘要】 大跨度桥梁的轻柔化以及结构形式与功能的日趋复杂化,加大了桥梁施工的难度,也对桥梁运营养护提出了更高的要求。一方面大跨度桥梁的建设是一个漫长而复杂的过程,其间要经过多次结构体系的转换,且易受到外界环境的影响,因此尚未成型的桥梁结构在施工期间的安全性能较成桥状态脆弱。若桥梁建设过程中使用的施工方法或工艺不当,将导致成桥时主梁的线型和内力分布严重不合理,进而加快桥梁的老化速度,降低桥梁的服务能力和使用寿命,严重者还易造成安全事故;另一方面,桥梁结构在正常运营过程中,由于环境侵蚀,车辆超载等因素的作用,将不可避免地出现各种损伤和病害,进而影响桥梁结构的使用性能和安全性能。因此,为了保证桥梁结构在施工和运营过程中的安全,必须从施工开始即对大跨度桥梁结构进行全寿命健康监测,包括传统意义上的施工控制、成桥荷载试验和运营健康监测。三者在监测内容、监测手段、技术路线上有相同之处,可共享监测设备及监测信息,故可将三者综合考虑,进行三位一体计,从而建立桥梁结构的全寿命健康监测系统。无论是从经济角度还是从系统和监测信息的完整性以及信息的衔接角度来看,建立全寿命健康监测系统都是最合理的。它可以实现对桥梁结构重要参数的全寿命周期内的跟踪监测,全面掌握当前桥梁结构的真实状态,并预测未来结构的行为与状态,因而能够尽早地发现和预报桥梁结构的损伤,及时采取必要地养护维修方法,从而杜绝或减少桥梁安全事故的发生。本文以峪道河斜拉桥为背景,综合运用有限元理论、静动力模型修正、模态参数识别、优化算法与结构仿真等理论和方法,对大跨度斜拉桥结构全寿命健康监测系统中的几个关键问题进行了深入和系统的研究,主要研究工作和成果包括如下几个方面:1.研究了大跨度斜拉桥施工过程中的索力、应力和挠度监测方法,完成了峪道河大桥的施工监控。通过对平行钢绞线斜拉索施工工艺的深入分析,建立了平行钢绞线斜拉索挂索、张拉过程中索-梁-塔力学模型,推导了钢绞线单根张拉时的初始拉力计算公式。对采用光纤光栅应变传感器进行应力测量时的温度补偿进行了研究,给出了埋入式光纤光栅温度补偿计算公式。在峪道河大桥的施工控制过程中,索力、应力和挠度实测值与理论预测值基本一致,证明了所提出的施工控制方法的有效性,保证了桥梁施工过程中的结构安全。2.提出了一种基于振动信号统计特征的损伤识别方法。首先推导了正弦激励和白噪声激励下结构的位移响应,进而得到各测点位移响应的统计特征(相关系数、回归系数和协方差之比)表达式。由于这些统计特征中包含了结构的刚度信息,故可通过其变化识别结果损伤产生的刚度下降,形成基于振动信号统计特征的损伤识别方法。与传统损伤识别方法相比,该方法只需对两测点位移响应进行简单的线性回归计算,无需进行复杂的频谱分析,且具有很好的抗噪性。三种常见桥梁结构(简支梁桥、连续梁桥和斜拉桥)损伤的数值模拟结果表明,基于振动信号统计特征的损伤识别方法能够对损伤位置、程度和范围进行有效的识别。3.提出了一种基于主梁挠度在线监测的损伤预警方法。不失一般性,假设桥梁结构为对称结构,以保证主梁跨中挠度的测量精度为目标,给出了确定倾角仪数量和布设位置的方法,即倾角仪数量为采用振型分解法计算跨中挠度所需用奇数阶振型数目的两倍,且关于跨中对称布置在奇数阶振型的倾角最大处。不同荷载形式下三种常见桥梁结构的数值模拟结果表明,该方法能够较为精确地测量主梁跨中挠度。另一方面,假设一段时间内过桥车辆的统计特征不变,则主梁挠度统计特征的变化仅取决于结构的刚度改变,故通过对在线监测所得的主梁挠度进行统计分析和假设检验,即可根据其统计特性的变化情况进行损伤预警。数值模拟结果表明,根据挠度统计特性的变化能够有效地进行结构损伤预警,但无法准确判断损伤位置和程度。4.提出了一种基于索力增量变化的损伤识别方法。假设斜拉桥主梁为弹性地基梁,可以证明,只有集中荷载作用附近区域内的斜拉索索力有明显改变,且损伤单元内力的释放等同于在该单元两端反向施加集中力,即结构损伤只引起损伤单元附近几个斜拉索索力的变化。因此,根据集中荷载作用下索力的变化情况可以确定损伤位置,然后借助于有限元模型修正技术进一步实现对损伤程度的识别。另一方面,将过桥车队简化为均布荷载加集中荷载,可以证明斜拉索最大索力回归系数仅决定于索力影响线面积之比,且单元损伤只影响其附近斜拉索的索力影响线面积。因此,可以通过正常运营过程中各索力回归系数的变化进行损伤位置判断,从而克服了在车载未知情况下无法估计索力大小的缺陷。数值模拟计算结果表明,通过同一荷载(单一重车)作用下的索力增量变化能够有效地识别主梁和斜拉索损伤,而由索力回归系数的变化只能识别斜拉索损伤。5.提出了面向养护策略的大跨度桥梁技术状态评估方法。养护策略包括养护时机和养护规模两项内容,养护时机由主要受力构件的最不利状态决定,而养护规模则由各构件技术状态的加权平均确定。本文提出的面向养护策略的大跨度桥梁技术状态评估方法采用各构件的建设费用作为其进行评估时的权重,从而避免了采用层次分析法确定构件权重时的主观随意性,使得不同的技术人员更易得到趋同的评估结果。此外,文中还以峪道河斜拉桥为例,详细介绍了使用该法进行桥梁技术状态评估的步骤。6.提出了一种由简单到复杂、由粗糙到精确的分层次桥梁承载能力评估方法。该方法通过检测或监测信息,估算或反算抗力钢筋的有效面积,进而实现对截面抗力的修正;将过桥车辆等效为滤过泊松过程,通过现场交通流量调查(车辆组成、车速和车流量),确定一段时期内车辆荷载效应最大值的分布,并根据95%的保证率下的荷载效应值与其设计值之比,确定更加符合桥梁实际运营情况的车辆荷载效应修正系数;提出了由设计可靠度到目标可靠度的荷载效应分项系数调整方法,并结合材料强度统计特性的现场检测信息,给出了现由真实可靠度到目标可靠度的荷载效应分项系数修正公式。该方法的本质就是在保证桥梁结构安全的前提下,充分挖掘桥梁结构的潜在承载能力,以避免不必要的浪费。7.根据施工控制与健康监测相结合、检测与监测相结合、实时监测与适时监测相结合以及健康监测与管理系统结合,注重实用性,兼顾经济性和科学研究的原则,设计了峪道河大桥健康监测系统,完成了各类传感器的布设,并介绍了健康监测软件的基本功能与实现方法。
【Abstract】 The design of large span bridge is increasingly flexible, meanwhile, structure form and function is increasingly complicated. As a result, the bridge construction and maintenance is more difficult. The construction of long span bridge is a long and complicated process. With times transformation of structure system, the bridge is vulnerable affected by environment. Bridge under construction is weaker than the bridge after completion. The improper construction method could cause serious unreasonable girder liner type and force distribution. Moreover, the bridge aging rate is accelerated and the service capacity with bridge life is deteriorated. Therefore, safety accidents will occur with serious cases. In order to ensure the safety condition of bridge under its construction and operation, comprehensive monitoring and testing is needed. The construction control, loading test and operation monitoring are the three stages for bridge health information acquisition. The contents, methods, and techniques are the similar in these three stages. In order to share the equipments and information, these stages are combined to the unified life cycle monitoring system. From the point of information integrity and economic, the bridge life cycle healthy monitoring is reasonable. This monitoring system can track the important parameters of the bridge to explore the bridge’s real condition and predict the future condition. The damage can be detected earlier, and maintenance can be done timely to prevent safety accident.Based on Yudaohe cable-stayed bridge, A deep study have been done on the structure damage detection and condition assessment using the Finite Element Method, dynamic model updating methods, modal parameters identification, optimal algorithm, and structural simulation. The main contents of this dissertation are described as follows:1. The monitoring method and analysis techniques of cable forces, stress, and deflection is studied. The construction of parallel strand cable is researched, and the mechanic model of construction is established. The initial tension of single strand is given. The temperature modification technique for FBG strain sensors is studied, and the modification formulation is given. The measurement value of cable forces, stress and deflection is consistent with the prediction value in the construction control of Yudaohe bride, and the bridge construction safety is assured.2. A structure damage detection method based on statistical characteristics of vibration signals is presented. Displacement responses under sine and white noise excitation are pursued firstly, and then the expression of displacement response statistical characteristics between different nodes is given. Because there is lots of structural stiffness information in the statistical characteristics, the feasibility of taking the statistical characteristics as a damage detection index has been proved in theory. Compared with the normal damage detection method, Spectrum Analysis with complicated calculation is avoided, and only regression analysis with simple calculation is needed. With another advantage of satisfactory noise immunity, this damage detection index is very suitable for on line structure healthy monitoring. Based on simulation for three normal types bridge (simple supported bridge, continue bridge and cable-stayed bridge), the identification efficiency of damage location, degree and extension has been proved.3.A structure damage early warning method based on girder deformations monitoring is presented. The girder deformations under the truck loads have been fitting with the vibration modals. The modal order for fitting is determined by the deformations contribution of various vibration modals. Taking the symmetry of normal bridges into account, the number and installation positions of the inclinometers have been presented for mid-span displacement. The inclinometer number cannot be less than twice of the odd modal order, and the symmetrical positions with the maximum angles on the odd vibration modals have been selected for installation. Displacement simulations for three type bridges under different class loads have calculated. The results show that the mid-span displacement can be efficiently fitting. The statistical characteristics of traffic are supposed unchanged in a period. So the change of main girder deformation is determined by the deterioration of structural stiffness. Statistic Analysis and Hypothesis Testing have been done to the main girder displacement data for damage promotion. Using different Hypothesis Testing methods for various bridges with different spans, simulation results show that statistic characteristic of mid-span displacement is sensitive to structure damages.4.A structure damage detection method based on line cable force monitoring is presented. As a typical cable supported structure, the cable forces of cable-stayed bridge contain lots of structure information. Therefore, the prerequisite for using cable forces to damage detection is sufficient. First, the main girder is equivalent to an elastic foundation beam, and only few cable forces near the load change under the concentrated load. The internal forces of the damage element are released, and cause the same effect of acting the concentrated forces on the node of damage element. So the conclusion that damage only caused few cable forces change nearby is given. With this conclusion, structural damage positions detection can be done based on the change of cable forces under the fixed truck load. Damage degree can be identified with Model Updating technique. The truck troop can be simplified to a uniform load with a concentrated load. The regression relationship between the cable forces can be found, and the regression relationship is determined by the areas of cable force influence lines. As the change of the influence line only appear nearby the damage element, so the damage position location identification can be finished with the cable forces regression ratios. The disadvantage that the rationality of cable forces is difficult to judge with the unknown truck load is avoided. Simulation has been done to prove the damage detection efficiency of cable forces regress ratios.5.A method of condition assessment of long span bridge based on management strategies is presented. Maintenance timing and scale are included in management strategies. The maintenance timing is determined by the worst condition of the main components, and the maintenance scale is determined by the weight average conditions of all components. In order to avoid individual subjectivity of calculating components weights by AHP method, the components construction costs is treated as the weights and different engineers can easily obtain the same result. An example of Yudaohe cable-stayed bridge is given to illustrate the procedures using this bridge condition assessment method.6. A bridge loading carrying capacity evaluation method with several rating levels is given to formulate a procedure from simple rough level to complex exact level. Section resistance, load effect and partier factors influence the bridge loading carrying capacity together. As the effective areas of reinforcement can be evaluated by inspection or monitoring information, so the section resistances can be modified. The truck rank is taken as a Filtered Poisson Process. With sited traffic investigation, the maximum distribution of load effects can be identified in a period. The load effect with 95% guaranteed rate is compared with its design value to formulate the modified factor. Partial factors are calculated from design reliability to target reliability. With more statistical properties information of the material strength, the modification from actual reliability to target reliability can be realized. The essential idea of this method is to exploit the potential capacity to avoid unnecessary waste with the promise of bridge safety.7. With the combination of bridge design, construction, inspection and maintenance codes, the principle of healthy monitoring system is considering the practical applicability, economical efficiency, and scientific research. The Yudaohe bridge healthy monitoring system is introduced. The contents include various sensors arrangement with corresponding acquisition equipments and basic functions of healthy monitoring software.
【Key words】 Long span Cable-stayed Bridge; Damage Detection; Model Updating; Bridge Assessment; Statistics Characteristics;