【作者】 李方明； 陈国兴；

【Author】 Li Fangming Chen Guoxing (Institute of Geotechnical Engineering,Nanjing University of Technology,Nanjing 210009,P.R.China)

【机构】 南京工业大学岩土工程研究所；

【摘要】 由提供的淮安市典型场地土的钻孔资料,采用建筑抗震设计规范法(规范法)、美国国家地震研究中心提供的简化方法(NCEER)和谢君斐—陈国兴方法(谢陈法)三种国内外著名的液化判别方法及 BP 神经网络液化判别模型方法对22个钻孔的120标贯点进行了液化判别,根据所有标贯点的液化判别结果可以看出:①四种地震液化势评估方法的计算结果基本一致;②NCEER 法比谢陈法计算的结果要保守;③谢陈法与规范法的计算结果很吻合,判别成功率相差不大;④规范法相对 BP 神经网络模型方法判别时要保守一点,BP 神经网络方法比其他三种方法更符合每个钻孔各标贯点的实际液化规律。最后根据判别结果结合整个地区的工程地质情况给出了该地区的初步液化综合评估结果,并给出了部分典型地质剖面土层地震液化势评价图。更多还原

【Abstract】 Based on the drill data of typical foundation soil in Huai’an city,BP neural network model for liquefac- tion estimation and three famous liquefaction potential estimation methods--construction anti-earthquake design criterion method,the methodology suggested by the workshop sponsored by the National center for Earthquake En- gineering Research of American (NCEER) and Xie Junfei-Chen Guoxing method—are applied to the seismic lique- faction estimation of 120 standard penetration tests of 22 drills.According the liquefaction estimation results of all standard penetration tests,it can show that:①the results of the above four seismic liquefaction potential evalua lion is almost identical,②the result with NCEER method is more conservative than that with Xie Junfei-Chen Guoxing method,③the result with Xie Junfei-Chen Guoxing method is close to that with criterion method,and the successful rates of two estimation methods are subtle,④the criterion estimation method is more conservative than BP neural network model for estimation,but BP neural network model is more identical with the practical liquefac- lion rule for every drill’s standard penetration tests than those other methods.According to the estimation results, the all-around preliminary liquefaction evaluation results in the area are obtained by analysis of the engineering ge- ology of the whole area,enclosing the seismic liquefaction potential evaluation chart of some typical geology.更多还原