基于纤维随机分布的单向复合材料拉伸破坏过程模拟

单向复合材料的极限拉伸强度受纤维强度变化和纤维随机分布的影响。本文提出了一种合理的纤维随机分布假设,并以此建立了剪滞数值模型。利用蒙特卡罗模拟方法结合现有模型分析了单向复合材料的拉伸失效过程,得到极限拉伸强度。与文献模型相比较的结果表明,本文的模型结果更理想。     

基于蒙特卡罗随机有限元方法的随机多孔介质内流体自然对流不确定性研究

为分析孔隙率不确定性对多孔介质方腔内自然对流换热的影响,发展了一种基于KL（Karhunen-Loeve展开）-蒙特卡罗随机有限元算法的随机多孔介质内自然对流不确定性分析数理模型及有限元数值模拟程序框架。通过K-L展开及基于拉丁抽样法生成多孔介质孔隙率随机实现,并耦合多孔介质自然对流有限元程序,进行随机多孔介质内自然对流传热数值模拟,得出了多孔介质内流场与温度场平均值与标准偏差,并分析了孔隙率不确定性条件下Da数对Nu数的影响。结果表明,孔隙率不确定性对多孔介质方腔内自然对流有重要影响。随机多孔介质内流场及温度场与确定性条件下的流场及温度场存在一定偏差,Nu数标准偏差随着Da的增大先增大后减小。     

基于S变换的非平稳随机过程演变功率谱密度估计

提出了一种基于S变换的估计Priestley非平稳随机过程演变功率谱密度的方法。此方法的根本在于,相对于S变换的\"变换核\",Priestley非平稳随机过程的调制函数为慢变函数。因此,非平稳随机过程的S变换可视为相位修正后的另一非平稳随机过程。推导出了对应于特定频率点的S变换瞬时均方值和非平稳随机过程演变功率谱密度之间的关系式。将功率谱密度函数表达为有限个频率点的级数展开,通过求解一组代数方程,就能得到级数展开中每个频率点的时变系数,由此,可给出非平稳随机过程的演变功率谱密度。由于级数展开中的高斯形状函数不依赖于时间,因此,本文所提算法具有较高的计算效率。最后,给出了均匀调制和非均匀调制非平稳随机过程演变功率谱估计的算例。     

移动随机荷载作用下梁桥非平稳随机振动的PEM-FFT方法

针对移动随机载荷作用下桥梁结构非平稳随机振动问题,建立了一种基于频域分析的虚拟激励-傅里叶变换方法(PEM-FFT)。与通常的非平稳随机振动的时频分析方法不同,提出的方法完全在频域上执行。其主要特色是能够给出随机输入与随机输出的频域关系,表明移动载荷作用下结构非平稳随机振动分析仅需对载荷的确定性移动函数项进行计算。数值算例考虑了等截面/变截面简支梁受移动随机载荷作用的问题,首先与通常的时频分析方法进行对比,验证了PEM-FFT方法的正确性和有效性,进一步讨论了不同结构形式以及不同载荷移动速度对响应演变统计量的影响。     

随机过程激励下随机结构系统可靠度分析的一种方法

提出了随机过程荷载激励下，具有随机参数的结构系统可靠度分析的一种方法，该方法基于首次超越破坏机制，分析随机过程荷载激励下，结构参数（随机变量）取某一确定向量时的条件失效概率，采用Ｍｏｎｔｅ Ｃａｒｌｏ技术模拟结构参数的随机性，由条件失效概率给出随机结构的无条件失效概率，最后对中方法和程序作了检验，并进行了实际计算。     

考虑平稳随机响应的显式结构拓扑优化

随机载荷是工程结构在服役中经常承受的一种复杂的载荷形式,通常采用统计学特性对其进行描述。对随机载荷作用下的结构进行拓扑优化设计是一项极具挑战性的工作,其主要难点在于,(1) 传统隐式拓扑优化方法的设计变量数巨大,且用于结构动态性能拓扑优化问题时存在虚假模态等数值不稳定问题; (2) 对结构的随机动力响应统计量及其灵敏度进行计算需要极大的计算量; (3) 隐式拓扑优化框架下的分析模型与优化模型强耦合,导致结构有限元模型具有极高的自由度,进一步加剧了上述困难。本文基于移动可变形组件框架和虚拟激励法理论,提出了一种平稳随机载荷作用下结构的显式拓扑优化设计方法。通过将一系列可移动和可变形的结构组件作为优化的基础单元,实现了使用少量设计变量描述结构拓扑构型的目的。采用虚拟激励法、自由度删除技术和模态位移法有效降低了对结构进行随机振动分析和灵敏度分析的计算量。在此基础上,以结构柔顺度的标准差为目标函数、以设计域内实体材料的体积为约束条件,实现了限带白噪声作用下结构的拓扑优化设计,并通过数值算例验证了本文方法的有效性。     

CFRP加固RC梁的模糊随机可靠度数值模拟

CFRP加固RC结构可靠度的不确定性,不仅要考虑随机性,也要考虑模糊性.本文根据模糊数学与可靠性理论,研编了引入模糊示性函数的Monte Carlo计算程序,通过试验讨论模糊示性函数的确定方法.利用Matlab软件,直接产生随机变量数组,解决Monte Carlo 抽样困难.CFRP加固梁破坏试验获得全部随机变量的统计参量.编制基于Matlab的Monte Carlo计算程序,完成了对CFRP加固RC梁抗弯模糊随机可靠度数值模拟,模拟计算结果得到传统FOSM法验证.     

随机有限元分析的二级区域分解并行求解算法

采用蒙特卡罗模拟(MCS)和加权积分法对二维问题进行随机有限元分析。尽管MCS方法对任何有确定解的问题都具有求解精度高的优点,但由于求解所需的计算量巨大使其应用受到限制。利用并行求解技术可有效地处理这种密集型计算问题。基于有限元分裂对接法(FETI)的并行特性并利用预处理共轭梯度法(PCG)的求解高效性,结合整体子区域实现(GSI-PCG)和FETI法,提出二级求解算法,并在工作站机群上实现了数值算例。算例计算结果表明本文GSI(PCG)-FETI算法具有较高的并行加速比和并行效率,具有良好的性能,可有效地进行二维问题的随机有限元分析。     

随机超载对疲劳裂纹扩展迟滞效应的模拟

考虑超载的迟滞效应,对随机超载作用下的疲劳裂纹扩展进行了模拟计算.载荷谱为在基本恒幅循环载荷基础上加入一以泊松流发生的随机超载序列,超载的大小为均匀分布.采用裂纹闭合模型考虑超载的迟滞效应,认为裂纹张开应力在超载引起的塑性区内按线性规律衰减.循环续循环模拟计算出裂纹从初始长度一直到疲劳破坏的扩展曲线.据此,计算了各种超载发生强度和大小下的疲劳裂纹扩展寿命的平均值与标准差.     

用虚拟激励法求解非比例阻尼线性体系的非平稳随机地震响应

应用复振型分解方法,将非比例阻尼线性体系在地震作用下的动力方程求解问题转化为若干个广义复振子的求解与叠加问题。通过假定地震地面运动为一零均值的非平稳随机激励,应用虚拟激励法原理,推导得到了广义复振子动力坐标计算的一般公式,进而得到了非比例阻尼线性体系非平稳随机地震响应计算的一般解答。由于可以选择少量共轭复振型的影响进行计算,对于大型复杂非比例阻尼结构,其随机地震响应计算工作量可以大幅度减小。算例证实了这种方法的可靠性及可行性。     

随机参数多自由度体系在平稳随机反应下的系统动力可靠性

对具有随机参数的多自由度体系,提出了求解其系统动力可靠度上、下限的一种计算方法。考虑结构的物理和几何参数具有随机性,从结构随机响应的频域表达式出发,利用求解随机变量数字特征的代数综合法和矩法,导出了随机参数多自由度体系在平稳随机激励下的平稳随机反应均方值的数字特征,再由动力可靠性的Poisson公式导出了随机参数结构的动力可靠度的计算公式,然后根据系统可靠性分析方法,分析了随机参数多自由度体系的系统动力可靠性,最后给出了系统动力可靠度上、下限的计算公式,并给出一个算例。     

Numerical Analysis of Stationary Dynamic Processes in Anisotropic Inhomogeneous Cylinders

Efficient numerical-analytical approaches are proposed for analyzing stationary dynamic processes. The approaches are used to study the propagation of elastic waves in anisotropic inhomogeneous cylinders with circular and noncircular cross sections and the natural vibrations of anisotropic solid and hollow cylinders of finite length with different end conditions. The dependence of the dynamic characteristics of cylinders on their geometrical and mechanical parameters is analyzed __________ Translated from Prikladnaya Mekhanika, Vol. 41, No. 8, pp. 3–40, August 2005.     

Markovian Diffusive Representation of Nonrational Distributed Random Processes and Application to Turbulence Simulation over Structures

Vehicles moving through the atmosphere or ocean are frequently subjectto distributed random excitations. This situation occurs for examplewhen turbulent boundary layers develop over their surfaces since thewall pressure fields are distributed random fields. It also occurs whenlarge-scale turbulent flows are encountered by the vehicles such asturbulent winds for an aircraft. These random fields may inducestructural vibrations, mechanical loads, fatigue and damage. They arefrequently described by spectral models (Corcos or Chase models forturbulent boundary layers, Dryden or Von Karman models for turbulentwinds) which involve nonrational terms.For such nonrational models, the development of a Markovian timesimulator relies on a rational approximation of the wavevector-frequencyspectrum obtained from an identification stage. This paper presents ageneral method to manage this stage, that provides us with a family ofstable rational approximations that converges towards the true model asthe dimension increases. These approximations are then used to simulatethe random field.We first give an exact but infinite dimensional state spacerepresentation of any nonrational distributed random field. It is basedon a diffusive formulation and the frequential decoupling property ofthe spatial Fourier transform. Then the discretization of this exactmodel leads to stable finite-dimensional approximations over aprescribed frequency range.The approach is applied to the simulation of a theoretical two-dimensional(2D) turbulent wind spectrum.     

线性随机桁架结构的平稳随机响应分析

研究了随机桁架结构的平稳随机响应问题。同时考虑结构的物理参数、几何尺寸的随机性，从结构平稳随机响应在频域上的表达式出发，利用求解随机变量函数矩的方法，导出了随机桁架结构在平稳随机激励下位移响应均方值的均值、均方差和变异系数的计算表达式。通过算例考察了随机荷载激励下结构物理参数、几何尺寸的随机性对结构随机响应的影响。     

随机结构在随机激励下的二阶摄动随机势能泛函及其应用

利用小参数摄动法,建立了随机结构在随机激励下的二阶振动随机势能泛函。并由此推导了二阶摄动随机变原理,作为应用,建立了随机有限元的计算列式。     

Stochastic stability of parametrically excited random systems

Multidegree-of-freedom dynamic systems subjected to parametric excitation are analyzed for stochastic stability. The variation of excitation intensity with time is described by the sum of a harmonic function and a stationary random process. The stability boundaries are determined by the stochastic averaging method. The effect of random parametric excitation on the stability of trivial solutions of systems of differential equations for the moments of phase variables is studied. It is assumed that the frequency of harmonic component falls within the region of combination resonances. Stability conditions for the first and second moments are obtained. It turns out that additional parametric excitation may have a stabilizing or destabilizing effect, depending on the values of certain parameters of random excitation. As an example, the stability of a beam in plane bending is analyzed.Published in Prikladnaya Mekhanika, Vol. 40, No. 10, pp. 135–144, October 2004.     

Long Asymptotic Correlation Time for Non-Linear Autonomous Itô's Stochastic Differential Equation

Itô's stochastic differential equations theory is a common approach to analysis of stochastic phenomena in various systems. In many applications, an important feature of the systems is the flicker effect. It is well known that it cannot be described with linear autonomous scalar equations of the above kind. The reason is that the flicker effect is usually associated with a correlation time which is much greater than the correlation time in the linear case. In the present work, we discuss modelling of the long correlation time with the help of non-linear autonomous scalar Itô's stochastic differential equation which includes non-linear drift. The expression for the asymptotic correlation time as time separation tends to zero is derived in terms of the equation. We formulate the condition for this time to be long in the above sense. It is pointed out that this condition can hold if the nonlinear damping is reduced compared to the linear case. These results are illustrated with an example of the equation with non-linear drift of a specific form.     

大跨度桥梁空间脉动风场的计算机模拟

针对现有Deodatis方法模拟大跨度桥梁空间脉动风场中存在的计算量问题,通过对谱分解矩阵引入插值近似,减少谱分解的次数,从而提高该谐波合成法的计算效率,并节省内存花费,实现了对三维空间脉动风场的有效模拟。改进方法模拟的脉动风速样本仍保持各态历经性,且逐渐收敛到目标功率谱。用改进的Deodatis方法模拟了润扬长江悬索桥桥面主梁上作用的纵向脉动风速。结果表明,该改进措施对Deodatis方法的应用效果非常明显,改进的Deodatis方法模拟脉动风速样本的相关函数与目标相关函数均吻合良好。尽管改进后的Deo-datis方法对谱分解矩阵采用了插值近似,但模拟的随机风速样本仍具有很好的精度。     

Construction of Invariant Torus Using Toeplitz Jacobian Matrices/Fast Fourier Transform Approach

The invariant torus is a very important case in the study of nonlinear autonomous systems governed by ordinary differential equations (ODEs). In this paper a new numerical method is provided to approximate the multi-periodic surface formed by an invariant torus by embedding the governing ODEs onto a set of partial differential equations (PDEs). A new characteristic approach to determine the stability of resultant periodic surface is also developed. A system with two strongly coupled van der Pol oscillators is taken as an illustrative example. The result shows that the Toeplitz Jacobian Matrix/Fast Fourier Transform (TJM/FFT) approach introduced previously is accurate and efficient in this application. The application of the method to normal multi-modes of nonlinear Euler beam is given in [1].     

Transient Stochastic Analysis of Biodegradable Contaminant Transport: First-Order Decay

A stochastic methodology was used to analyze the field-scale transport of solutes in heterogeneous aquifers with first-order biodegradation. Spectral methods and perturbation techniques were utilized to develop expressions for the field-scale effective parameters in the mean transport model. Expressions were obtained for the longitudinal and transverse macrodispersivity coefficients, and effective velocity and an effective decay parameter for statistically anisotropic, and isotropic, heterogeneous porous medium, respectively. The behavior of these parameters was described as function of time and log K correlation scale. The expressions for asymptotic values of the dispersivity coefficients and effective velocity and decay parameters were also derived for the isotropic case.