【作者】 潘峰；

【导师】 吕大刚；

【作者基本信息】 哈尔滨工业大学 ， 岩土工程， 2007， 硕士

【摘要】 地震动(Earthquake Motion)、地震需求(Seimsic Demand)和抗震能力(Seimsic Capacity)是“基于性能的地震工程”(Performance-Based Earthquake Engineering, PBEE)理论的三块基石。工程结构的地震需求与地震动参数之间的关系称为地震需求模型,由于地震发生的随机性和地震动的随机过程性,以及工程结构自身参数的随机性,因此结构的地震需求模型本质是随机的,研究结构地震需求与地震动参数之间概率关系的学科称为“概率地震需求分析(Probabilsitic Seismic Demand Analysis, PSDA)”,相应的关系称为“概率地震需求模型( Probabilsitic Seismic Demand Model, PSDM)”。工程结构的概率地震需求分析是结构抗震可靠度分析、结构地震易损性分析以及结构地震风险分析的基础,同时也是采用全概率方法进行结构概率抗震性能设计和概率抗震性能评定的重要组成部分;另外,概率地震需求分析的结果也可以为地震损失估计和防震减灾决策提供科学的依据。因此,对重大土木工程结构和基础设施系统进行概率地震需求分析,不仅具有重要的理论意义,而且具有重要的工程实用价值。本文在美国太平洋地震工程研究中心(Pacific Earthquake Engineering Research Center, PEER)提出的新一代“基于性能的地震工程”概率决策框架之下,分别以地震动的峰值参数(PGA、PGV、PGD)和谱参数(Sa、Sv、Sd)作为概率地震需求分析的输入变量,以结构的最大层间位移角、最大顶点位移角和等效单自由度位移延性系数作为整体地震需求参数,选取不同震中距和震级的100条实际地震纪录,以基于非线性纤维梁—柱单元的动力时程分析为工具,从只考虑地震动变异性与同时考虑结构参数变异性两个角度出发,分别建立钢筋混凝土框架结构基于对数线性回归模型和响应面模型的概率地震需求模型,结合场地的地震危险性,对结构的地震需求危险性和地震需求易损性进行系统深入的分析。研究结果表明:1.只考虑地震动变异性的地震需求模型与同时考虑地震动变异性和结构参数变异性的地震需求模型相比,分析结果偏于保守。2.地震动峰值参数与谱参数相比,回归分析的离散程度较大。3.通过使用对数线性回归模型和响应面模型来建立概率需求模型,发现分别得到的危险性曲线和易损性曲线有一定差异,说明概率地震需求分析依赖于对其选择的模型。更多还原

【Abstract】 Earthquake motion, seismic demand and seismic capacity are three cornerstones of Performance-Based Earthquake Engineering (PBEE) theory. The relationships between seismic demand parameters of engineering structures and earthquake motion intensity measures are called seismic semand models. Since the seismic activity is random in nature, and the earthquake motion is a stochastic process; what’s more, the structural model and its parameters are random due to both aletory and epistemic uncertainities, therefore, the structural seismic semand model is also random in nature. Probabilistic Seismic Demand Analysis (PSDA) is the subject which sets up the probabilistic relationships of seismic demand parameters and earthquake motion intensity measures, and the results of PSDA is named Probabilistic Seismic Demand Models (PSDM).Probabilistic seismic demand analysis of engineering structures is the foundation of structural seismic reliability analysis, structural seismic fragility analysis and structural seismic risk analysis; meanwhile, it also is an important part of probabilistic performance-based seismic design and assessment of structures by using the full probability (level-3) method. On the other hand, the results of PSDA can also provide scientific basis for earthquake loss estimation and earthquake disaster mitigation decision-making. Therefore, it is very important to apply PSDA in civil engineering structures and infrastructure systems.Under the probabilistic decision-making framework of the new generation of PBEE proposed by Pacific Earthquake Engineering Research Center (PEER), this study uses the peak value parameters (PGA, PGV and PGD) and spectrum parameters (Sa, Sv and Sd) of earthquake motion intensity measures as the input variables of PSDA, and takes the Inter-Storey Displacement Angle (ISDA), Roof Displacement Angle (RDA) and equivalent single-degree of freedom displacement ductile coefficient (μd) as the global seismic demand parameters of structures. 100 actual seismic motion records are selected according to different epicentral distances and magnitudes, and then the dynamic time history analysis based on nonlinear fibre beam-column elements is used as the computational tool for PSDA. From the two viepoints of only considering earthquake motion variations and simultaneously considering structural parameter variations as well as earthquake motion variations, two PSDMs of the reinforced concrete frame structures, namely logarithm linear regression model and response surface model, are set up. Combined with Probabilistic Seismic Hazard Analysis (PSHA) results of the site, the hazard and fragility of the global seismic demand are analyzed thoroughly and systematically.The main conclusions of this thesis are as follows:1. Compared with the case of simultaneously considering structural parameter variations and earthquake motion variations, the case of only considering earthquake motion variations is more conservative than the former.2. Compared with the results of PSDA based on peak value parameters of earthquake motion, the results PSDA based on spectrum parameters is better.3. It is found that threre is a certain difference in PSDMs based on logarithm linear regression model and response surface model. It is indicated by the hazard curves and the fragility curves that PSDA depends on PSDMs.更多还原