静力触探参数自相关距离确定方法与影响因素分析

发布时间：2018-05-06 00:39

本文选题：自相关距离 + 波动函数法　；参考：《岩土力学》2017年S1期

【摘要】：在岩土工程可靠性分析中,土性参数空间变异性的分析至关重要,其中土体的固有变异性可以通过随机场理论获取方差折减函数和自相关距离值进行分析。基于天津临港静力触探试验(CPT)得到的锥尖阻力,对波动函数法确定自相关距离过程中所遇到的问题进行深入研究,并分析了去趋势化方法、取样间距、取样范围等影响自相关距离的因素。结合空间递推平均法的改进方法,首次对应用波动函数法确定自相关距离时出现的多个峰值的选择开展分析,结果表明对于所分析场地的地层静力触探数据而言,应选取波动函数曲线中第一个峰值作为该土层的自相关距离值。通过去趋势化后的随机分量统计特性及对应自相关距离计算结果的对比发现,去趋势化方法要依实际情况而定,并不是所有情况都可以使用线性去趋势化方法,建议最高选用二次非线性拟合结果对原始数据去趋势化;当取样范围较大,即触探孔数量较多时自相关距离随取样间距的增大而增大。分析还表明,对于同样的样本间距,随着触探孔数量的增加,自相关距离是逐渐减小的,但这种差距随着取样间距的增大而逐渐缩小。
[Abstract]:In the reliability analysis of geotechnical engineering, the spatial variability analysis of soil parameters is very important, and the inherent variability of soil can be analyzed by random field theory with variance reduction function and autocorrelation distance. Based on the cone tip resistance obtained from the static penetration test (CPT) of Tianjin Port, the problems encountered in the determination of autocorrelation distance by the wave function method are studied in depth, and the detrend method and sampling spacing are analyzed. Sampling range and other factors affecting autocorrelation distance. Combined with the improved spatial recursive averaging method, the selection of multiple peak values when the wave function method is applied to determine the autocorrelation distance is analyzed for the first time. The results show that for the static sounding data of the analyzed site, The first peak value of the wave function curve should be chosen as the autocorrelation distance of the soil layer. By comparing the statistical characteristics of random components and the corresponding autocorrelation distance, it is found that the method of de-trending depends on the actual situation, and that not all cases can be characterized by linear de-trending. It is suggested that the maximum quadratic nonlinear fitting result should be used to detrend the original data, and the autocorrelation distance will increase with the increase of sampling distance when the sampling range is large, that is, the number of penetration holes is more. The analysis also shows that for the same sample spacing, the autocorrelation distance decreases with the increase of the number of penetration holes, but the gap decreases with the increase of the sampling spacing.
【作者单位】：中国科学院武汉岩土力学研究所岩土力学与工程国家重点实验室;天津城建大学天津市软土特性与工程环境重点实验室;厦门轨道交通集团有限公司;
【基金】：国家自然科学基金项目(No.11672320,No.41372291)~~
【分类号】：TU43

【参考文献】