• 全国中文核心期刊
  • 中国科技核心期刊
  • 美国工程索引(EI)收录期刊
  • Scopus数据库收录期刊
高广运, 石超, 陈青生. 地震荷载等效循环周数的预测模型[J]. 岩土工程学报, 2015, 37(11): 2040-2044. DOI: 10.11779/CJGE201511014
引用本文: 高广运, 石超, 陈青生. 地震荷载等效循环周数的预测模型[J]. 岩土工程学报, 2015, 37(11): 2040-2044. DOI: 10.11779/CJGE201511014
GAO Guang-yun, SHI Chao, CHEN Qing-sheng. A predictive model on equivalent number of strain cycles for earthquake loads[J]. Chinese Journal of Geotechnical Engineering, 2015, 37(11): 2040-2044. DOI: 10.11779/CJGE201511014
Citation: GAO Guang-yun, SHI Chao, CHEN Qing-sheng. A predictive model on equivalent number of strain cycles for earthquake loads[J]. Chinese Journal of Geotechnical Engineering, 2015, 37(11): 2040-2044. DOI: 10.11779/CJGE201511014

地震荷载等效循环周数的预测模型

A predictive model on equivalent number of strain cycles for earthquake loads

  • 摘要: 荷载等效循环周数的计算在分析地震动力响应中有重要作用。目前等效循环周数的计算方法大多是建立在累计损伤理论的基础上,这些方法一般需要已知地震时程曲线,而预测分析时并不知道实际的地震时程曲线,且等效转换处理较麻烦。为了解决以上问题,统计了涵盖不同震中距级别和震级级别的296条水平地震记录,计算出作用于相对密实度为45%的砂土的等效循环周数。基于统计分析的非线性混合效应模型,建立了单向及双向地震荷载等效循环周数与震级和震中距之间的预测模型。采用回归分析中常用的值检验法及残差分布方法对预测模型进行检验,结果表明预测误差在合理范围内。通过实测值与预测值比较,证明了预测模型在统计意义上的可行性与准确性。

     

    Abstract: The equivalent number of strain cycles plays an important role in seismic response analyses. The existing calculation methods for the equivalent number of strain cycles are mostly based on accumulated damage theories. However, the time history of an earthquake load is required in the prediction using these methods, which is complicated in the equivalent conversion. In order to address these issues, 296 records of horizontal seismic time-history records, which cover different kinds of the epicenter distance and magnitude, are used in this study. The equivalent number of strain cycles is calculated for the sand with a relative density of 45% subjected to the selected earthquake load. Based on the statistical analysis using the nonlinear mixed-effects model, a predictive model for the equivalent number of strain cycles is proposed for both unidirectional and bi-directional earthquake loads. By examining the p-value and the residual distribution, the results show that the predictive values are reasonable. The proposed predictive model is validated by the measured data.

     

/

返回文章
返回