论文部分内容阅读
通过日本KiK-net台网数据库的各种地震记录,研究了土体的非线性行为对场地响应的影响。这个台网由不少于688个地表—井下仪器构成,从这个台网还获得了直到井孔深度的剪切波和压缩波速度剖面的特征。为了通过计算每个场地的地表与井下频谱比来表征土体的线性行为,我们挑选了井下台站地震动峰值加速度(PGA)<10cm/s2的事件。使用强震动事件(PGA>50cm/s2)计算的场地响应曲线相对线性表征的变化被认为是由非线性土体行为造成的。为了描述每个事件土体非线性行为对场地响应的影响,我们提出了场地响应曲线相对线性估计的变化(放大或衰减)百分比(PNL ev,非线性百分比)和有关的位移频率(Sh ev)。这些参数被用于估计非线性场地响应显著区别于相应线性场地响应的概率。我们发现,不管何种场地,这个概率即使对低输入地震动峰值加速度(井下传感器的值不小于30cm/s2)也是重要的。这表明,在中强地震动场地响应评估中必须考虑非线性土体行为。此外,对记录了至少两个强震事件(井下PGA>50cm/s2)的KiK-net台网数据库的54个场地,我们定义了每个场地的表征土体非线性行为对场地响应影响的另外4个参数:(1)地震动峰值加速度阈值(PGA th),定义为PNL ev高于10%的地震动峰值加速度值;(2)地震动峰值加速度为50cm/s2的特定场地的PNL(PNL site);(3)地震动峰值加速度为50cm/s2的卓越频率的特定场地移位(Sh site);(4)在非线性和线性场地响应之间观测到衰减的频率(fNL)。我们观察到在fNL之下的频率,非线性土体行为可增强放大作用。我们发现fNL介于场地响应基频率和卓越共振频率之间,近地表具有V S反差的场地触发低地震动峰值加速度输入阈值的非线性行为。这些结果表明,非线性行为多发生在地表土层中。此外,通过研究地表的地震水平与垂直频谱比的土体非线性行为,我们发现它们对评估fNL频率和位移频率(Sh site)能够给出满意的结果(等价于井下场地响应的分析),这表明本研究获得的部分结果可以扩展到没有井下传感器的其他数据库。
Through the various seismograms recorded in Japan’s KiK-net database, the effect of soil nonlinear behavior on site response was studied. The network consists of no less than 688 surface-subsurface instruments, from which shear-wave and compressional-wave velocity profiles are obtained up to the borehole depth. In order to characterize the linear behavior of the soil by calculating the ratio of the subsurface to the subsurface at each site, we selected events with a peak ground acceleration (PGA) of <10 cm / s2 at the downhole station. The change in the relative linear characterization of the site response curve calculated using a strong shock event (PGA> 50 cm / s2) is considered to be caused by nonlinear soil behavior. In order to describe the effect of soil nonlinear behavior on site response for each event, we present the change (magnification or attenuation) percentage (PNL ev, non-linearity) and relative displacement frequency (Sh ev) of the relative linear estimate of the site response curve . These parameters are used to estimate the probability that the non-linear site response differs significantly from the corresponding linear site response. We found that this probability is important even for low input ground motion peak accelerations (downhole sensors have a value of not less than 30 cm / s2) regardless of the site. This indicates that the nonlinear soil behavior must be considered in the response assessment of moderately strong ground motions. In addition, for the 54 sites in the KiK-net database with at least two strong earthquakes (PGA> 50 cm / s2), we define the effect of each site’s characterization on soil response to site response 4 parameters: (1) PGA th, which is defined as the peak value of ground motion with PNL ev higher than 10%; (2) PNL with specific peak ground motion acceleration of 50 cm / s 2 site; (3) a specific site shift at an excellent frequency of 50 cm / s2 for peak ground motions; and (4) a frequency of fNL attenuation observed between nonlinear and linear site responses. We observe the frequency below fNL and nonlinear soil behavior enhances amplification. We found that the fNL is between the fundamental response frequency and the excellent resonance frequency of the site, and the non-linear behavior of the site with V S contrast near the surface triggers the input threshold of the peak acceleration of low ground motion. These results show that non-linear behavior occurs mostly in the surface soil. In addition, we found that they can give satisfactory results (equivalent to the analysis of the downhole site response) for evaluating the fNL frequency and the Sh site, by studying the nonlinear behavior of the earth above and below the vertical seismic spectrum. This shows that some of the results obtained in this study can be extended to other databases without downhole sensors.