【作者】 艾庆华；

【导师】 李宏男； 王东升；

【作者基本信息】 大连理工大学 ， 防灾减灾工程及防护工程， 2008， 博士

【摘要】 近几十年的桥梁震害促进了桥梁抗震技术的快速发展,世界上主要桥梁抗震规范已采用延性设计,基于性态的抗震设计方法成为目前的研究热点和发展趋势,对桥梁体系非线性地震反应和破坏机理的深入了解和把握是性态设计的关键所在。本文首先详细总结和分析该领域的主要文献,其次从数值分析角度研究了利用简化塑性铰模型和弹塑性纤维梁柱单元,计算钢筋混凝土桥墩整体和局部地震损伤性态指标或参数的精度及可行性,最后从模型振动台试验角度研究和比较了基于现行规范设计、基于位移设计和可能发生弯剪破坏模式的钢筋混凝土短柱桥墩的抗震性能及破坏机理。论文在以下几个方面有所创新或有新的认识:1.开发了钢筋混凝土桥墩拟静力试验数值模拟与地震损伤评价程序DLUT-RC主要功能是进行钢筋混桥墩拟静力试验滞回性能的数值分析和利用真实试验数据及数值模拟结果进行桥墩抗震性态指标的计算。该程序包括两种分析模型:简化塑性铰模型和弹塑性纤维梁柱单元模型。塑性铰模型可考虑不同试验装置导致的P-△效应。纤维单元模型包括刚度法和柔度法单元,可考虑材料非线性、几何非线性和节点锚固钢筋粘结滑移影响。程序采用了较为精确反映钢筋、混凝土反复荷载下力学性能的Chang-Mander本构关系模型。以力学模型计算等效塑性铰长度,以Lehman基于应变的试验统计公式计算节点纵筋粘结滑移,建议了一种新的简化塑性铰模型。2.研究了钢筋混凝土桥墩地震反应数值模型与抗震性态评价指标的匹配关系利用简化塑性铰模型和纤维单元模型,主要计算了Lehman等完成的5根不同剪跨比及配筋率的钢筋混凝土桥墩的截面曲率、残余位移、混凝土和钢筋应变幅值、纵筋低周疲劳损伤指数等损伤评价指标,并与试验结果比较。认为不同模型计算的滞回曲线、残余位移可以和试验有很好吻合;塑性铰模型计算的极限曲率、纵筋应变幅值和试验结果较为接近,但对于剪跨比λ≥8的高墩,极限曲率计算值小于试验值,将导致偏于危险的结果:纤维单元模型,通过合理划分单元网格或确定合适的高斯积分点数,对截面曲率、纵筋应变幅值刚度法单元计算结果较为接近试验值,而柔度法单元明显大于刚度法单元。柔度法单元计算的纵筋疲劳指数要大于刚度法单元,但两者的统计离散性都很大。所有模型都很大程度低估了桥墩破坏时的混凝土压应变幅值,纤维模型要略好于简化塑性铰模型。此外,就目前常用的数值分析模型及其能力而言,认为混凝土压应变和纵筋低周疲劳损伤指数尚不适合作为钢筋混凝土桥墩的抗震性态评价指标。3.进行了延性弯曲破坏和脆性弯剪破坏的钢筋混凝土桥墩地震模拟振动台试验分别基于规范设计方法、基于位移设计方法以及考虑配箍率不足和剪跨比较小情况设计了一组(8根)钢筋混凝土桥墩试件,通过振动台试验从破坏形态、延性、耗能等多个方面评价比较了它们的抗震性能。表明按课题组建议的基于位移方法设计与按现行规范方法设计的桥墩二者抗震性能相当,但前者可明确预知(预设)桥墩在不同地震水准下的抗震性态,且在增加约束箍筋条件下可一定程度减少纵向钢筋使用量,能够达到预期的延性抗震要求;短柱桥墩在地震作用下反应位移延性系数较大,而耗能偏小,更易于破坏。更多还原

【Abstract】 Investigations of recent earthquake disasters develope bridge seismic technologies, as a result, the main bridge seismic design codes in the world have adopted ductility design method. Performance based bridge seismic design method is current research focus, of which understanding of nonlinear seismic response and damage mechanism of bridge system is the key component. At first, the main documentations in this field are summarized and analyzed in this dissertation. Then feasibility and precision for simplified plastic hinge model and inelastic fiber beam-column element to calculate integral and local seismic damage performance parameters are studied. At last a shaking table test is designed and implementd to understand and compare the seismic performance and damage mechanism of RC bridge pier specimens designed by current seismic code, by displacement based seismic design method or likely to occur flexure-shear failure.The main work of the study are:1. A computer program for numerical simulation and seismic damage evaluation of RC bridge piers subjected to cyclic loading—DLUT-RC is developed. The main functions of the program are hysteretic analysis of RC bridge piers and calculation of seismic performance indices from experimental or computing results. Two kinds of analysis models are employed in the program: simplified plastic hinge model and inelastic fiber beam-column element. P-A effects caused by variant test devices are considered in plastic hinge model. Fiber element includes stiffness-based element and flexibility-based element, in which material nonlinearity, geometry nonlinearity and bond slip effect of anchoring steel in joint are accounted for. Chang-Mander models for steel and concrete strain-stress relationship are employed, which reflect material mechanical property subjected to cyclic loading precisely. A new plastic hinge model is suggested, in which mechanics-based effective plastic hinge length is adopted and an experiment static formula based on strain suggested by Lehman is used to account for bond slip of steel in joint.2. The matching between seismic response analysis models and seismic performance indices for RC bridge piers is studied. Simplified plastic hinge model and fiber element are employed to compute damage evaluation indices of 5 bridge pier specimens by Lehman, such as section curvature, residual displacement, strain amplitude of steel and concrete, low cycle fatigue damage index of longitudinal steel etc. Comparison between experimental and simulating results is made, and it is concluded that the computed hysteretic curves and residual displacements by both models fit experiment results very well. With regard to plastic hinge model, limit curvature and longitudinal steel strain amplitude approximate experiment data, but the limit curvature is less than experiment whenλ≥8, which causes a dangerous design. With regard to fiber element model, section curvature and longitudinal steel strain amplitude calculated with stiffness-based element fit the experiment very well by proper dividing element. Low cycle fatigue damage indices of longitudinal steel computed by flexibility-based element are greater than that by stiffness-based element. Both models underestimate concrete compressive strain amplitudes when bridge piers fail, but fiber element is a little better than simplified plastic hinge model. According to the study of the dissertation, compressive strain of concrete and low cycle fatigue damage index of longitudinal steel are unfit for seismic performance indices of reinforced concrete bridge piers.3. Shaking table tests for RC bridge piers with ductile flexural damage model and with brittle flexure-shear damage model are studied. 8 RC bridge pier specimens are designed and fabricated by current seismic design code, by displacement-based seismic design method, and with low lateral reinforcement ratio and with low shear span ratio respectively for shaking table test. Damage states, displacement ductility and energy dissipation are compared to access their seismic performance. It is concluded that the specimens designed by displacement-based seismic design method have the equivalent seismic performance with those designed by current seismic design code, but the former can anticipate the seismic performance of bridge piers, reduce longitudinal steel amount by increasing lateral reinforcement and approach anticipated ductility. It is also shown that displacement ductility response of low shear span bridge piers is hige, and energy dissipation is low, which should be paid much attention to in bridge seismic design.更多还原