Practical second‐order reliability analysis applied to foundation engineering

A practical and efficient approach of implementing second‐order reliability method (SORM) is presented and illustrated for cases related to foundation engineering involving explicit and implicit limit state functions. The proposed SORM procedure is based on an approximating paraboloid fitted to the limit state surface in the neighborhood of the design point and can be easily carried out in a spreadsheet. Complex mathematical operations are relegated to relatively simple user‐created functions. The failure probability is calculated automatically based on the reliability index and principal curvatures of the limit state surface using established closed‐form SORM formulas. Four common foundation engineering examples are analyzed using the proposed method and discussed: immediate settlement of a flexible rectangular foundation, bearing capacity of a shallow footing, axial capacity of a vertical single pile, and deflection of a pile under lateral load. Comparisons with Monte Carlo simulations are made. In the case of the laterally loaded pile, the friction angle of the soil is represented as a one‐dimensional random field, and pile deflections are computed based on finite element analysis on a stand‐alone computer package. The implicit limit state function is approximated via the response surface method using two quadratic models. Copyright © 2011 John Wiley & Sons, Ltd.     

Variational elastic solution for axially loaded piles in multilayered soil

Most analytical or semi‐analytical solutions of the problem of load‐settlement response of axially loaded piles are based on the assumption of zero radial displacement. These solutions also are only applicable to piles embedded in either a homogeneous or a Gibson soil deposit. In reality, soil deposits consist of multiple soil layers with different properties, and displacements in the radial direction within the soil deposit are not zero when the pile is loaded axially. In this paper, we present a load‐settlement analysis applicable to a pile with circular cross section installed in multilayered elastic soil that accounts for both vertical and radial soil displacements. The analysis follows from the solution of the differential equations governing the displacements of the pile–soil system obtained using variational principles. The input parameters needed for the analysis are the pile geometry and the elastic constants of the soil and pile. We compare the results from the present analysis with those of an analytical solution that considers only vertical soil displacements. The analysis presented in this paper also provides useful insights into the displacement and strain fields around axially loaded piles. Copyright © 2011 John Wiley & Sons, Ltd.     

Discussion and improvement on Monte Carlo method and geometric method

After careful analyses of the popular Monte Carlo method and geometric method in reliability analysis, the misuse of Monte Carlo method with several non-uniform distribution random variables is pointed out. As a result, a correct procedure sampling procedure of Monte Carlo method is presented. Meanwhile, in order to satisfy the needs of practice in reliability analysis, an extended geometrical optimized method(EGOM) is proposed, its procedure under conditions of correlation and non-normal distribution variable is also presented. Furthermore, several examples from others’ reliability analysis practice with the proposed EGOM prove its correctness and effectiveness.     

Reliability updating in geotechnical engineering including spatial variability of soil

At geotechnical sites, deformation measurements are routinely performed during the construction process. In this paper, it is shown how information from such measurements can be utilized to update the reliability estimate of the geotechnical site at future construction stages. A recently proposed method for Bayesian updating of the reliability is successfully applied in conjunction with a stochastic nonlinear geotechnical finite element model. Therein, uncertainty in the soil material properties is modelled by non-Gaussian random fields. The structural reliability evaluations required for the Bayesian updating are carried out by means of subset simulation, an efficient adaptive Monte Carlo method. The approach is demonstrated through an application to a sheet pile wall at a deformation-sensitive geotechnical construction site.     

基于响应面法的边坡稳定二阶可靠度分析

提出了基于响应面法的边坡稳定二阶可靠度分析的实用算法。选择U空间中的随机变量,通过空间变换和相关矩阵分解,计算试验点的功能函数;通过迭代算法构造响应面、以确保通过最小的计算量获得最优精度,并在此基础上进行FORM/SORM计算。以一岩石边坡的平面滑动问题为例,通过与蒙特卡洛模拟、FORM及随机响应面法的比较,证明了该方法的准确性和高效性。分析了参数的相关性及试验点取值范围对计算结果的影响,讨论了可靠度分析结果中参数敏感性和物理属性问题。该方法可为实际边坡问题的可靠度分析提供参考,并可以用来进行基于可靠度分析的加固设计。     

层状地基中轴向受荷单桩的边界元法分析

以层状地基内部作用一竖向集中力时的广义Mindlin解作为边界单元法的基本解,对层状地基中的轴向受荷单桩进行了分析,对基本解的奇异性处理方法进行了改进。考虑了桩的可压缩性和长径比对桩-土荷载传递规律和沉降特性的影响,编制了计算程序,并进行了数值分析和计算。结果表明,该方法具有较快的计算速度和良好的计算精度。     

基于认知聚类分区方法的边坡可靠度分析

提出了边坡可靠度分析的一种新的全局优化方法--认知聚类分区方法。该方法主要包括5个步骤：分区、随机抽样、计算极径L、回代、计算可靠指标及验算点。给出了相应的计算流程图,并编写了基于C语言的计算程序KCPREL。最后,以岩质边坡稳定可靠度问题为例证明了所提方法的有效性。结果表明,认知聚类分区方法能同时计算出可靠指标和验算点,并能获得全局最优解。该方法的计算精度和蒙特卡洛模拟方法相当,计算效率远远高于传统的蒙特卡洛模拟方法。此外,该方法在分析含有复杂的隐式及非线性功能函数的边坡稳定可靠度问题方面体现出明显的优越性。等步长认知聚类分区方法能全面且均匀地搜索角度,从而得到更准确的验算点。为了保证足够的计算精度及减小计算量,建议步长取10º以内。     

Probabilistic analysis of strip footings based on enhanced Kriging metamodeling

Two advanced Kriging metamodeling techniques were used to compute the failure probability of geotechnical structures involving spatially varying soil properties. These methods are based on a Kriging metamodel combined with a global sensitivity analysis that is called in literature Global Sensitivity Analysis-enhanced Surrogate (GSAS) modeling for reliability analysis. The GSAS methodology may be used in combination with either the Monte Carlo simulation (MCS) or importance sampling (IS) method. The resulting Kriging metamodeling techniques are called GSAS-MCS or GSAS-IS. The objective of these techniques is to reduce the number of calls of the mechanical model as compared with the classical Kriging-based metamodeling techniques (called AK-MCS and AK-IS) combining Kriging with MCS or IS. The soil uncertain parameters were assumed as non-Gaussian random fields. EOLE methodology was used to discretize these random fields. The mechanical models were based on numerical simulations. Some probabilistic numerical results are presented and discussed.     

钻孔灌注桩单桩承载力的可靠度研究

采用无量纲随机变量的极限状态方程,结合天津市钻孔灌注桩试桩资料进行可靠度分析。采用了JC法和Monte Carlo法两种方法进行可靠指标的计算,并对影响可靠指标的各种因素进行了分析。还计算了单桩承载力的抗力分项系数。     

Implementation of Bayesian theory on LRFD of axially loaded driven piles

In this paper, a framework based on Bayesian theory and proof pile load test results was used to update resistance factors of axially loaded driven piles. Prior to implementation of the framework, resistance factors were calibrated based on the distribution of the measured-to-predicted pile ultimate bearing capacity using the results of static pile load tests conducted to failure. These resistance factors and the distribution were considered to be “prior.” The prior distribution of the measured-to-predicted ultimate bearing capacity was updated based on Bayesian theory to incorporate additional proof pile load test results. Using the measured-to-predicted load distributions and the updated (or posterior) measured-to-predicted bearing capacity distributions, resistance factors were calibrated (or updated) from the first-order reliability method (FORM) for two different target reliability indices, 2.33 and 3.0. This research attempted to use the results of proof pile load tests, which are generally conducted to verify pile designs, to update resistance factors. The updated resistance factors varied substantially depending on the proof pile load test results. Therefore, the Bayesian implementation can contribute to economical pile designs.     

Boundary element analysis of axially loaded piles embedded in a multi-layered soil

In a field, piles are likely installed in a multi-layered soil. Analysis of axially loaded piles in a multi-layered soil is complicated and deserves more attention. A boundary element method is used in this study to analyze an axially loaded single pile in a multi-layered soil using the solution for vertical and horizontal axisymmetric ring loads in a multi-layered elastic medium. Good and reasonable agreement is obtained between the proposed and published solutions for a single pile in a homogenous soil, a finite soil, and a Gibson soil. The proposed solution is also used to evaluate an axially loaded single pile in a multi-layered (8 layers) soil.     

列车荷载下冻土路基非平稳随机振动分析的时域显式法北大核心CSCD

随着我国冻土地区铁路运营里程的不断提升,冻土路基在随机列车荷载作用下的动力响应分析成为了急需解决的工程问题。本文以青藏铁路某路基横断面为例,采用时域显式蒙特卡罗模拟法,计算其在随机列车荷载作用下响应的统计特征。首先,提出列车驶过施加在路基顶面荷载的快速计算方法,并引入轨道不平顺及列车行驶速度两类随机参数,以生成随机分析所需的大量荷载样本。然后,采用数值积分方法将运动方程在时域上进行离散,建立任意离散时刻冻土路基动力响应关于列车荷载的显式表达式。基于该显式表达式,可以高效地实施蒙特卡罗模拟,得到在随机列车荷载作用下路基关键响应的均值、标准差和峰值等统计量。采用该方法,分析了路基在夏冬两个季节不同深度处的随机动力响应,发现在夏季路基浅层中的位移和速度响应最为剧烈。数值算例表明,时域显式蒙特卡罗模拟法在冻土路基随机振动分析中具有理想的计算精度和计算效率。     

基于刚性承台群桩沉降的可靠度研究

针对在总结目前的群桩可靠度设计理论完全是基于承载力,提出了对于超大超长群桩基础采用基于沉降的可靠度设计概念,其中沉降计算模式考虑了桩身压缩以及桩端刺入量。在讨论了计算参数的分布规律后,用蒙特卡罗方法进行了可靠度指标的计算。通过一个算例表明,荷载水平增加会导致群桩可靠度指标急剧降低;可接受位移增加,可靠度指标增大,但这种趋势逐渐变缓。     

SVM在地下工程可靠性分析中的应用

将支持向量机应用到地下工程可靠性分析中,通过将支持向量机分别与一阶二次矩和蒙特卡洛结合,提出了基于支持向量机的可靠性分析方法,利用数值模拟构造学习样本,通过支持向量机学习,建立变形与随机变量之间映射关系的支持向量机表达,进而实现隧道极限状态函数及其偏导数的显式表达,从而计算隧道的可靠性指标。该方法避免了传统可靠性分析的缺点。算例分析结果表明,该方法计算效率高、结果可靠,对含有大量随机变量的复杂岩土工程可靠性分析具有很大的潜力,具有广泛的应用前景和工程价值。     

非均质地基中群桩竖向荷载沉降关系分析

运用剪切位移法计算了桩轴向荷载传递因子。对于桩端采用线性的荷载传递函数,推导了基于弹塑性模型的单桩竖向荷载沉降的解析解。分析过程中考虑了土体强度沿深度线性变化的特性和桩土间的滑移现象,因此更符合大部分土体的实际性状。在此基础上,建立了考虑桩土滑移的桩-桩相互作用系数的计算公式,并将上述方法应用于群桩的分析,获得了群桩的荷载沉降特性。该分析方法克服了目前应用较多的弹性理论方法夸大桩土相互作用的缺点,单桩和群桩的荷载沉降曲线的分析结果和实测数据吻合,证明了该方法的合理性。     

Prediction of the shaft resistance of nondisplacement piles in sand

Using pile segment analysis, the mobilized shaft resistance of axially loaded nondisplacement piles in sand is investigated here. It is accepted that the shaft capacity of piles constructed in granular soils is highly influenced by the mechanical behavior of soil–structure interfaces forming adjacent the piles skin. Adopting the thin interface layer as a load transfer mechanism, a simple but accurate critical state compatible interface constitutive model is introduced. After evaluation, the interface model in conjunction with the pile segment analysis is applied for the prediction of the shaft resistance mobilized in nondisplacement piles. The proposed approach takes into account the influences of pile diameter and surface roughness together with the effects of the surrounding soil density and stiffness on the mobilized shaft resistance. The performance of the proposed method is verified by comparing its predictions with the experimental data of various model piles covering wide ranges of length, diameter, roughness, and surrounding soil properties. Copyright © 2012 John Wiley & Sons, Ltd.     

基于克里金模型的边坡稳定可靠度分析方法

边坡工程的复杂性不仅表现为各岩土参数的变异性和非确定性,而且还在于其极限状态功能函数的非解析性及隐式性。以Janbu 法为例,研究了隐式功能函数下易于执行的边坡工程稳定可靠度计算方法。首先,调用边坡极限平衡模式获得岩土基本参数,并利用拉丁超立方试验设计抽取影响边坡稳定性基本参数的适量初始样本。其次,采用地质统计学中的克里金（Kriging）各向异性关联映射方法,将边坡功能函数值表述为随机过程。然后,结合主动学习方法,基于搜索规则调整训练样本,通过反复迭代循环确定满足实际工程精度的随机过程所表示的边坡功能函数。最后,调用随机过程函数通过验算点法（JC法）获得边坡的失效概率。工程算例分析表明,文中方法的求解精度与蒙特卡洛模拟方法相当,但计算过程简明,效率高,更具工程实用性。     

评估灌注桩极限承载力的概率能量法

本文研究用概率能量方法评估灌注桩极限承载力的可行性,研究所用数据取自南京市区的２００根桩例的ＣＡＰＷＡＰ反分析与静载试验结果。根据能量守恒定律并结合动态桩－土相互作用模型,详细地推导了桩极限承载力计算公式。并对其中的模型参数,Ｊｓ,Ｊｔ,Ｑｓ,Ｑｔ作为随机变量处理,统计分析了它们的平均值,标准差以及概率分布函数。用蒙特卡洛法模拟确定了桩极限承载力的概率分布函数,并进一步提出了计算桩极限承载力超过某一设计值的概率公式。通过计算的桩极限承载力与静载结果的良好可比性,说明了此方法的台理性与准确性。