【作者】 彭磊；

【导师】 谢礼立；

【作者基本信息】 中国地震局工程力学研究所 ， 防灾减灾工程及防护工程， 2011， 硕士

【摘要】 随着强震观测技术不断的发展,以及强震台网建设的规模不断扩大,强震记录的不断积累,特别是大量近断层强震记录的获得,人们开始观察到近场地震动完全不同于远场地震动的特点,具体就是各种所谓的近断层特殊效应。近断层地震动的特殊效应包括近断层地震动的集中性、地表破裂和永久位移、破裂的方向性效应、近断层的速度大脉冲、上盘效应。这些近断层地震动特殊效应使得断层周围的地震动展现出不同于以往远场地震动的性质,例如使得地震动在某一个区域变得很大,而其他区域相对较弱,或是地震动在长周期增大很多,短周期变化不大,甚至有的地震动有突然像脉冲一样的短暂大幅增大。这样我们有必要对这些近场效应进行详细研究,并希望在新修改的衰减关系中对近场地震动的预测能够包括这些近场效应对地震动造成的影响,这样就可以更加准确的预测近断层地区的地震动。汶川地震是我国近百年来发生在人口密集地区的最大的一次地震,造成了巨大的工程破坏和人员伤亡。但是源于这几十年来我国强震观测台站的迅速发展,在这次地震中,取得了我国强震观测史上最多的强震观测资料,特别是有一批近场的强震记录,使得我们详细研究本次地震的各个方面成为可能。文本着重研究了汶川地震的上盘效应。主要研究内容和结论如下:1、选取汶川地震近场区域的地震记录,研究并获得了汶川地震的衰减关系。并在此基础上得到所选取地震记录的残差。利用经典的残差分析方法分析上下盘地震动残差的分布,从而确定汶川地震中存在上盘效应。2、在确定汶川地震中存在上盘效应后,本文采用模型模拟上盘效应,从而定量的研究上盘效应。本文研究了前人提出来的两个上盘模型AS1996和KS2003,分别用于模拟北岭地震和集集地震中的上盘效应。发现这两个上盘模型都存在一些问题。具体是,(一)模型过于复杂,使用起来很不方便;(二)模型的主观随意性强烈,导致大量的不确定性。3、在确定了前人提出的模型中的缺点后。本文提出了一个新的模拟上盘效应的模型。新模型可以克服前两个模型的缺点。新模型使用简单方便,减小甚至消除主观性,可以使用所有可用数据,使得误差减小到最低限度。经过将新模型用于模拟北岭地震和集集地震中的上盘效应,并与AS1996和KS2003的模拟结果比较,新模型优势明显。4、利用新模型模拟了汶川地震中的上盘效应,具体模拟了水平地震动加速度峰值PGAH和水平加速度反应谱SAH的上盘效应。同时把模拟结果与NGA(Next Generation Attenuation Relationship)项目中的Abrahamson& Silva一组的衰减关系中的上盘因子AS NGA对汶川地震的上盘效应预测结果进行比较。发现汶川地震中的上盘效应比AS NGA的预测结果小,同时相比AS NGA的预测结果,汶川地震中的上盘效应在较远距离时仍然存在,而AS NGA预测结果显示在这个距离时,已不存在上盘效应。更多还原

【Abstract】 Along with the increasingly development of strong earthquake observation technology, increasingly extended construction of network station for strong earthquake observation, accumulation of earthquake records especially of those from near fault area, people started to find out the absolute different performance of the near fault ground motions in comparison with ground motions in far field. Near fault ground motions possess many special characteristics, such as rupture directivity, hanging wall effect, concentration of ground motions, surface rupture and flint-step, large velocity pulse. These characteristics cause the near fault ground motions to behave differently, other than what supposedly similar to the far field ground motions. For example, in the near fault area, ground motions in some areas are strong while mild in other areas, or ground motions in long periods are fierce while mild in short periods, or even ground motions have pulse like waves. So, it is necessary to study comprehensively these characteristics or effects of near fault ground motions in order to revise the old attenuation relationships to incorporate influence of these effects on ground motions. The revised the attenuation relationships are supposed to predict near fault ground motions more accurately.Wenchuan earthquake is the biggest one happened in densely populated area in our country. It caused such huge damage and heavy casualty. However, due to the rapid construction of strong-motion instrument in recent years, a large set of data has been obtained in this earthquake. This is the biggest set of strong ground motion data ever since, and enable an in-depth research into the aspects of this earthquake. This study focuses on the hanging wall effect of Wenchuan earthquake, and the main work includes:1. Selected near fault data of Wenchuan earthquake and data were processed, collected and classified. Based on the selected data, the Wenchuan earthquake specific attenuation relationship was regressed. Then, residuals of chosen data were obtained by subtracting attenuation relationship from original data. The hanging wall effect in Wenchuan earthquake was identified after the distribution of residuals on hanging wall and footwall was analyzed.2. After the recognition of hanging wall effect in Wenchuan earthquake, it is necessary to model the hanging wall effect for quantitative analysis of the hanging wall effect in the earthquake. There are two models AS1996 and KS2003 as denoted in this study put forward by previous researchers to the hanging wall effect in Northridge earthquake and Chi-Chi earthquake respectively. However, after a comprehensive analysis of the two models, some drawbacks are identified with the models. Specifically, firstly, the two models are complex and used with inconvenience; the other, the two models introduced large portion of subjectivity, to lead to big uncertainty as a result.3. After the recognition of the drawbacks of previous models, a new model was put forward in this study to avoid all problems with previous models. So, the new model or proposed one has advantages over previous models. It is easy to use and can reduce or even completely avoid subjectivity during the process of determination of models. Finally, the new model makes full use of all data, and error can be reduced as low as possible. The new model therefore was applied to model the hanging wall effect in Northridge earthquake and Chi-Chi earthquake, results of which were compared with each that of AS1996 and KS2003. Satisfying results were obtained as the comparisons show.4. The new model was applied to model hanging wall effect in Wenchuan earthquake, including PGAH and SAH. Meanwhile, the modeling results were drawn in comparison with the AS NGA’s prediction. It showed that AS NGA predicted pretty high level of hanging wall effect in Wenchuan earthquake. Besides, it failed also to predict the existence of hanging wall effect in a little bit far distance still.更多还原