【作者】 周大兴；

【导师】 闫维明； 陈彦江；

【作者基本信息】 北京工业大学 ， 土木工程， 2012， 博士

【摘要】 地震工程实践表明，土—结构相互作用对结构的抗震性能有重要的影响，是桥梁震害的一个重要原因，而减震控制则是减轻结构地震反应的有效手段。有鉴于此，本文对考虑土—结构相互作用的大跨径连续梁桥抗震性能和减震控制问题进行了较为系统深入的研究，主要工作和成果包括如下几个方面。（1）考虑土—结构相互作用连续梁体系地震响应的参数影响分析。采用直接法在FLAC-3D中建立了桥梁和土体的整体分析模型。对比分析了土层的密度、泊松比、剪切波速和厚度，地震动加速度峰值以及结构刚度对连续梁桥地震响应的影响，并总结了相关规律。根据研究结果可知，若桥址所在地为II类、III类、IV类场地或者局部冲刷线以下5倍桩径深度内存在较大范围的剪切波速不大于200m/s的土层，且设防烈度不低于7度时，连续梁桥（D类桥梁除外）的抗震分析应考虑土-结构相互作用的影响。（2）考虑土—结构相互作用三种结构体系抗震性能的对比研究。在Midas/Civil中采用改良的Penzien模型建立了连续梁桥三种结构体系（连续梁、刚构连续组合和连续刚构）的简化分析模型。从高矮墩、墩形和行波效应三个方面对比了这三种结构体系的抗震性能。通过分析可知，在某一地震响应方面，高矮墩和墩形对不同结构体系的影响不同。同一结构体系中，高矮墩和墩形对不同地震响应的影响也不同。行波效应对三种结构体系均有较大影响。（3）考虑土—结构相互作用连续梁体系减震控制研究。虽然粘滞阻尼器具有较好的减震效果，但其设计、制作等技术难度大，价格昂贵，而且养护成本和技术要求高，这些因素可能导致其在地震作用过程中不能发挥应有的作用。鉴于此原因，本文对大行程板式铅阻尼器进行了研究。在有限元和试验结果的基础上，提出了一种计算阻尼力的简化方法，以便于阻尼器的设计。从减震控制的分析结果来看，大行程板式铅阻尼器和粘滞阻尼器一样，也可以与支座并联作为分离型减震装置使用，且减震效果明显。由于具有较高性价比，大行程板式铅阻尼器在桥梁工程中具有很好的应用前景。另外，大跨径连续梁桥的减震控制应该考虑土-结构相互作用的影响，否则不仅在支座的设计上偏于不安全，而且可能会导致桥墩的普通钢筋用量偏大。（4）考虑土—结构相互作用大跨径连续梁桥的振动台子结构试验探索。通过试验发现，三种结构体系的试验结果与数值分析结果的规律基本一致。不过，界面力的获取方式、时滞、噪声影响以及振动台自身的控制精度等因素造成期望指令与实际指令是存在一定差异的，而且这种差异具有随机性。因此，试验结果与数值结果在大小上差别较大。（5）神经网络在振动台子结构试验中的应用研究。振动台子结构试验的一个重要研究内容是试验系统的时滞与补偿。时滞会影响试验的精度、系统的稳定，甚至造成结果的发散。在现代控制理论中，神经网络是解决时滞问题的一个有效工具。在振动台子结构试验中，可以利用神经网络对作动器的信号进行预测。通过仿真分析发现，神经网络的时滞补偿效果显著。更多还原

【Abstract】 From earthquake engineering practice, soil and structure interaction has much in-fluence on seismic performance of structure, and it is an important cause of bridgedamages, while damping control is an effective means to reduce the seismic re-sponse of structure. For the above reasons, the seismic performance and dampingcontrol of long-span continuous girder bridges have been studied, which are listed inthe following.(1) Parameter analysis of the seismic response of the continuous girder systemconsidering soil and structure interaction. With the help of FLAC-3D, the wholeanalysis model that includes soil and the bridge has been built using the directmethod. The influence of soil density, Poisson’s ratio, the velocity of soil shear wave,the thickness and type of soil and so on is summarized on SSI. According to the re-sults, if the location of the bridge is Class II, III or IV site, and the fortification in-tensity is not less than7, seismic analysis of continuous girder bridges (except forbridges belonging to Class D) should consider SSI.(2) Comparative study of the seismic performance of the three structure systemconsidering soil and structure interaction. With the help of Midas/Civil, the simpli-fied analysis model has been built using the improved Penzien model. And Seismicperformance of three structure systems has been studied from the difference of pierheight, pier type and traveling wave effect. According to the results, in a seismic re-sponse, the pier height and pier type have different influence with different structuresystems. And in a structure system, the pier height and the pier type have differentinfluence with different seismic response. Traveling wave effect has much influenceon the three structure systems: in lognitudinal direction, support displacement oftransition pier will increase, and other seismic response will basically reduce. Intransverse direction, seismic response may increase. So, doing research on the threestructure systems’ seismic performance should consider traveling wave effect.(3) Damping control of the seismic response of the continuous girder system con-sidering soil and structure interaction. Viscous damper has good damping effect, butits design and production are difficult and maintenance cost is high. And because ofthese factors, it can’t play a role in the process of earthquake. For this reason, thelarge stroke lead damper has been studied in this article. A simplified method ofcalculating the damping force has been proposed on the basis of the finite elementanalysis and test results, in order to facilitate the design of the damper. Like viscous damper, large stroke lead damper can also be placed with the bearingin parallel as a shock absorber used in long-span continuous girder bridges, and thedamping effect is good. With higher performance price ratio, large stroke leaddamper has good prospect in bridge engineering. In damping system，if consideringSSI, support displacement and the main pier deformation will increase, and the in-ternal forces at the bottom of the side pier and the main pier will be reduced. So, thedamping system should consider the influence of SSI, otherwise it is unsafe in thedesign of the bearing, and the amount of the ordinary reinforced bar in the side piermay increase.(4) Exploration of real-time substructure testing with shaking table for long-spancontinuous girder bridges considering soil and structure interaction. With the testingtechniques of real-time substructure testing with shaking table, the shaking table testconsidering SSI can be achieved. According to experimental data, the law in the re-sults of the three structure systems is basically the same with that in the results ofnumerical analysis. But the difference in value is large, and among the seismic re-sponse, the biggest difference is about20%in support displacement and the mainpier deformation. This is mainly related to the way to get interface force, time delay,noise, the control accuracy of shaking table and so on. In other word, when com-mand signal calculated by the numerical substructure drive the array by the controlsystem of shaking table, there is some difference between the table signal and com-mand signal, but this difference is random.(5) Application of neural network in real-time substructure testing with shakingtable. Studying on time-delay compensation is an important aspect in real-time sub-structure testing with shaking table. Time delay will affect the accuracy and the sta-bility of the system, and even cause divergence. In modern control theory, neuralnetwork is the effective way to compensate time delay. So, it can be used to predictthe signal applied to the actuator of the shaking table. According to the simulation,the effect of compensation is significant.更多还原