时程分析考虑高阶振型影响的多频段地震波选择方法研究

MULTI-BAND MATCHING METHOD FOR SELECTION OF GROUND MOTIONS IN TIME-HISTORY ANALYSIS CONSIDERING HIGHER MODES EFFECTS

针对结构抗震时程分析输入地震波选择问题,以加速度反应谱(或设计谱)为目标谱,在反应谱平台段和结构基本周期附近误差双控指标中,引入了由归一化振型参与系数确定的前几阶振型的权重系数,提出了考虑多频段加权匹配的时程分析输入地震波选择方法,以充分考虑高阶振型的影响。以美国SAC Steel Project提出的9层和20层抗弯钢框架Benchmark结构为实例,为方便将建议选波方法与较多地震波输入时程分析结果进行比较,以简单地震信息初选的小型地震波数据库中的20条地震波的均值(放大系数)谱作为目标谱,以这20条地震波时程分析得到的结构最大基底剪力和层间位移角均值为目标反应,前者考虑弹性状态,后者对应弹塑性情况。将按该文方法选出的3条波输入下结构反应均值与目标反应进行了比较分析,初步证实了建议选波方法的可行性。将该文方法与ASCE 7-05地震波输入要求方法对比,在弹塑性反应状态下,两者对结构最大层间位移角沿楼层分布规律的估计以及薄弱层位置的判断均与目标反应较为一致。ASCE 7-05方法输入下结构反应均大于目标反应,偏于安全。而按该文方法输入存在高估和低估结构反应的情况,建议附加"在0.2T1~1.5T1(T1为结构基本周期)区间内,3条地震波的平均谱应在目标谱上方"这一限制条件,以防止低估结构反应的情况发生。

For the problem of selecting ground motions in the aseismic time-history analysis of structures, this paper presents a weighted multi-band matching method, which can consider the significant contributions of higher modes. To adopt this proposed method, the acceleration response spectrum or the design spectrum is taken as the target spectrum, and two error indexes are adopted to check the matching degree between the response spectrum of each selected ground motion and the target spectrum within the flat range zone and the first some natural periods zones of structures. Note that the error index for checking latter zones is proposed to be calculated by introducing some weighting factors determined by modal mass participation factors. Then, the ground motions with smallest error indexes can be selected to calculate structural responses. in this paper, The 9-and 20-story Benchmark steel moment resisting frame structures proposed by the American SAC Steel Project are taken as examples. To conveniently compare the calculation results obtained by the limited ground motions selected according to the proposed method and a large number of ground motions, the target spectrum is determined as the average response spectrum of 20 earthquake records, which are originally selected just on the basis of earthquake information. The mean values of the maximum base shear and inter-story drift ratio calculated from 20 earthquake records are taken as the target seismic responses, the former corresponds to the seismic elastic response, and the latter is for the inelastic response. A total of 3 ground motion records are selected based on the proposed method, and the average responses obtained from the selected 3 records are compared with the target seismic responses. The agreement of their results preliminarily verifies the feasibility of the proposed method. Moreover, the proposed method is compared with the requirements of ground motion input in ASCE7-05, and it turns out that the structural seismic responses obtained by the proposed method are consistent with that based on ASCE7-05 when it is used to evaluate the distribution of structural maximum inter-story drift ratios and to recognize the weakest floors. The structural responses calculated according to ASCE7-05 are safer for their larger values than the target responses. Since both the overestimated and underestimated structural responses can be found by the proposed method, the proposed method is recommended to meet the additional condition that ＂within the range of 0.2 T1 and 1.5 T1, the average spectrum of the 3 ground motions should be displayed above the target spectrum＂, so that the underestimation of structural responses could be prevented.