PREDICTION OF SINGLE PILE SETTLEMENT BASED ON INVERSE ANALYSIS

PREDICTION OF SINGLE PILE SETTLEMENT BASED ON INVERSE ANALYSIS
复制标题

基于反演分析的单桩沉降预测

DOI:
10.3208/sandf1972.33.2_126
复制
发表时间:
1993
影响因子:
3.7
通讯作者:
Liu Wen
Liu Wen
中科院分区:
工程技术3区
文献类型:
--
作者:
Y. Honjo;Benny Limanhadi;Liu Wen

文献摘要

被引文献

相似文献

利用Poulos和Davis(1980)提出的桩沉降模型估计层状地基的杨氏模量,采用了一种以非线性回归方法为基础的逆分析技术。这是一个纯弹性模型,因此不考虑桩土之间的滑移。该方法不仅可以量化平均值,而且可以量化其中的不确定度。此外,还提出了一种基于一阶二阶矩(FOSM)的预测方法。分析了曼谷地区两组钻孔灌注桩在不同加载阶段的桩顶沉降和桩身荷载分布情况。研究发现,在不同地质环境和不同桩型的桩载试验结果的基础上,对参数(即各层的杨氏模量)进行估计,可以得到更可靠的估计。但是,预测结果可能不能直接反映估计结果的准确性,即如果地质条件和桩的配置与估计中使用的情况相差不大,则可以得到较好的预测结果。另一方面,如果这些条件非常不同,则可能导致错误的预测。所有这些结果都是基于具有多重共线性的桩载试验数据结构来解释的。
An inverse analysis technique which is firmly based on non-linear regression method has been adopted to estimate Young's modulus of layered ground using the pile settlement model proposed by Poulos and Davis (1980). This is purely an elastic model, therefore no slip between soil and pile could be taken into account. In this method, not only mean value but also uncertainty involved in it can be quantified. Furthermore, a prediction procedure based on the First Order Second Moment (FOSM) method is also proposed. Two sets of bored pile loading test data consisting of pile top settlement and load distribution along pile shafts at various loading stages from the Bangkok area are analysed. It is found that more reliable estimation of parameters is possible if they (i.e. Young's modulus of each layer) are estimated based on several pile loading test results which have been conducted in a variety of geological settings and pile configurations. However, the prediction may not directly reflect accuracy of the estimated results, i.e. good prediction can be obtained if the geological conditions and the pile configurations are not very much different from the ones used in the estimation. On the other hand, erroneous prediction could result if these conditions are very much different. All these outcomes are explained based on data structure of pile loading tests which exhibit the multicollinearity.