基于多重多级动力子结构的瞬态分析方法

提出利用多重多级子结构技术与Newmark算法求解结构动力学方程的高精度算法.该算法利用静凝聚技术列式简单,在凝聚过程中并不引入任何近似的优点,采用子结构周游树技术,分别对每个子结构求解Newmark等效平衡方程,最后通过回代求解得到整体结构的响应.由于该算法考虑了子结构内部自由度对整体求解的贡献,算法实施不受子结构划分方式的限制,因此可以得到系统高阶模态对响应分析的影响.该算法计算精度与传统的全结构求解相当,计算效率高,消耗计算机资源少,且可构造为统一的多重多级子结构综合分析算法框架.数值算例验证了该算法的正确性和有效性.     

求解接触问题的一种新的实验误差法

提出了一种带松弛因子的UZAW算法求解实验误差法中给定状态下的位移和接触力满足的等式方程，并证明了该算法是R超线性收敛的。整个区域被划分为多个子区域，不同子区域位移场的求解是独立的。还提出了一种带参数的以不完全因子分解为基础的预条件子共轭梯度法求解不同子区域位移场，该算法在块体规模较大时更加有效。     

开放式多体系统动力学仿真算法软件研发(I)DAEs求解算法构架设计

基于开放式工程与科学计算集成化软件平台SiPESC,研发了用于多体系统动力学时程分析的一类通用求解算法构架。该构架的核心思想是算法与数据相分离,整个构架由五个基本类及子类组成。本文重点阐述基本类的抽象过程,利用插件技术设计求解器的构架,进一步应用该构架实现了Newmark方法,HHT(Hilber-Hughes-Taylor)方法,Generalized α方法,Bathe方法及祖冲之类Symplectic方法等微分-代数方程组(DAEs)求解器的开发。研究工作表明,本文所提出的DAEs求解算法构架对多体系统动力学的时程分析具有良好的开放性和通用性,可方便进行各种新的DAEs求解算法的动态扩展。     

结构非线性动力分析混合时间积分并行算法

综合隐式和显式时间积分技术,对结构非线性动力反应分析提出一种并行混合时间积分算法.该算法采用区域分解技术.将并发性引入到算法中,即利用显式时间积分技术进行界面节点积分而利用隐式算法求解局部子区域.为实现并行混合时间积分算法,设计了灵活的并行数据信息流.编写了该算法的程序,在工作站机群实现了数值算例,验证了算法的精度和性能.计算结果表明该算法具有良好的并行性能,优于隐式算法.     

求解非线性方程组的混合遗传算法

非线性方程组的求解是数值计算领域中最困难的问题。大多数的数值求解算法例如牛顿法的收敛性和性能特征在很大程度上依赖于初始点。但是对于很多非线性方程组,选择好的初始点是一件非常困难的事情。本文结合遗传算法和经典算法的优点,提出了一种用于求解非线性方程组的混合遗传算法。该混合算法充分发挥了遗传算法的群体搜索和全局收敛性,有效地克服了经典算法的初始点敏感问题;同时在遗传算法中引入经典算法(Powell法、拟牛顿迭代法)作局部搜索,克服了遗传算法收敛速度慢和精度差的缺点。选择了几个典型非线性方程组,从收敛可靠性、计算成本和适用性等指标分析对比了不同算法。计算结果表明所设计的混合遗传算法有着可靠的收敛性和较高的收敛速度和精度,是求解非线性方程组的一种成功算法。     

开放式多体系统动力学仿真算法软件研发(Ⅰ)DAEs求解算法构架设计

基于开放式工程与科学计算集成化软件平台SiPESC,研发了用于多体系统动力学时程分析的一类通用求解算法构架。该构架的核心思想是算法与数据相分离,整个构架由五个基本类及子类组成。本文重点阐述基本类的抽象过程,利用插件技术设计求解器的构架,进一步应用该构架实现了Newmark方法,HHT(HilberHughes-Taylor)方法,Generalizedα方法,Bathe方法及祖冲之类Symplectic方法等微分-代数方程组(DAEs)求解器的开发。研究工作表明,本文所提出的DAEs求解算法构架对多体系统动力学的时程分析具有良好的开放性和通用性,可方便进行各种新的DAEs求解算法的动态扩展。     

????????????????????????????

合隐式和显式时间积分技术,对结构非线性动力反应分析提出一种并行混合时间积分算 法. 该算法采用区域分解技术. 将并发性引入到算法中,即利用显式时间积分技术进行界面 节点积分而利用隐式算法求解局部子区域. 为实现并行混合时间积分算法,设计了灵活的并 行数据信息流. 编写了该算法的程序,在工作站机群实现了数值算例,验证了算法的精度和 性能. 计算结果表明该算法具有良好的并行性能,优于隐式算法.     

非均匀介质中流动和传热问题的有限分析法

数值求解非均匀介质中的输运问题广泛应用于科学计算和工程领域．介质的强非均匀性给相关问题的准确求解带来极大的困难．近年来，本课题组将有限分析法拓展到该领域，建立了非均匀介质中输运问题的有限分析法．该算法基于网格奇点邻域内类拉普拉斯方程局部解析解构建，算法具有很高的精度，且不依赖于介质的非均匀性强度．不管相邻网格传导率差异如何，仅需对原始网格进行很少地细分就可以获得非常准确的计算结果，因此与其他传统数值算法相比，可以大幅提高计算精度和效率．该算法可广泛应用于求解非均匀多孔介质中的渗流、复合材料中的热传导及电场分布等问题．     

大型边界元方程组的并行直接分块求解算法

针对大型边界元方程组和网络微机机群环境提出了一种并行直接分块求解算法，算法基于分块高斯-若当消去法的原理，采用内外存交互技术，并行分块消去方法，节点超行的卷帘存储方案和并行环状循环逐次修正策略，增大了解题规模，提高了计算速度。算例计算结果表明该算法具有较高的并行加速比和并行效率，适用于大型问题的边界元法求解。     

一种改进的2n因子乘积形式的逆矩阵求解算法

复杂结构的有限元分析需要耗费巨大的内存和计算机时间。在此过程中，有限元线性方程组的求解时间占很大比例，因此，发展高效的线性方程组求解算法是提高有限元分析效率的关键，针对矩阵求逆问题，该文将有限元方程组的系数矩阵视为等分块矩阵，并基于等分块矩阵的概念推广了传统的逆矩阵的n因子和2n因子乘积形式。推广后得到的这组乘积形式在适当条件下又分别可以退化到传统的逆矩阵的n因子2n因子乘积形式，应用基于推广后得到的这组乘积形式的求逆算法来求解板材冲压成形有限元数值模拟中的大型有限元线性方程组，结果表明，该算法可以显著地提高大型有限元线性方程组的求解效率。     

Comparison of interfacial heat transfer coefficient estimated by two different techniques during solidification of cylindrical aluminum alloy casting

The interfacial heat transfer coefficient (IHTC) is necessary for accurate simulation of the casting process. In this study, a cylindrical geometry is selected for the determination of the IHTC between aluminum alloy casting and the surrounding sand mold. The mold surface heat flux and temperature are estimated by two inverse heat conduction techniques, namely Beck’s algorithm and control volume technique. The instantaneous cast and mold temperatures are measured experimentally and these values are used in the theoretical investigations. In the control volume technique, partial differential heat conduction equation is reduced to ordinary differential equations in time, which are then solved sequentially. In Beck’s method, solution algorithm is developed under the function specification method to solve the inverse heat conduction equations. The IHTC was determined from the surface heat flux and the mold surface temperature by both the techniques and the results are compared.     

Stabilization of Unstable Limit Cycles in Systems with Limited Controllability: Expanding the Basin of Convergence of OGY-Type Controllers

The underlying geometric structure of the standard OGY control scheme is analyzed. Some of the main mechanisms that under certain conditions lead to failure of the control algorithm are revealed. The limited controllability available in a system is investigated and it is shown that it may lead to serious problems that will significantly enlarge the state space region of failure of the standard OGY controller. A minimal distance algorithm is analyzed and shown to be, for some problems, more advantageous than the standard OGY technique. Nevertheless, for a broad category of problems, the minimal distance scheme is also shown to fail. As a solution for these problems, two new techniques are proposed: the penalized minimal distance and the multi-step OGY-type scheme. The standard OGY and minimal distance algorithms are particular cases of the new techniques proposed. Finally, we give a necessary condition that estimates the region of controllability under the multi-step OGY-type control. We demonstrate a significantly improved basin of convergence for the new multi-step OGY-type algorithm.     

An adaptive genetic algorithm for solving bilevel linear programming problem

Bilevel linear programming,which consists of the objective functions of the upper level and lower level,is a useful tool for modeling decentralized decision problems. Various methods are proposed for solving this problem.Of all the algorithms,the ge- netic algorithm is an alternative to conventional approaches to find the solution of the bilevel linear programming.In this paper,we describe an adaptive genetic algorithm for solving the bilevel linear programming problem to overcome the difficulty of determining the probabilities of crossover and mutation.In addition,some techniques are adopted not only to deal with the difficulty that most of the chromosomes may be infeasible in solving constrained optimization problem with genetic algorithm but also to improve the efficiency of the algorithm.The performance of this proposed algorithm is illustrated by the examples from references.     

Three-component micro-PIV using the continuity equation and a comparison of the performance with that of stereoscopic measurements

Recently, a number of techniques have been presented for the determination of the third “out-of-plane” velocity component in micro particle image velocimetry (micro-PIV) data. In particular, the conventional macroscopic stereo-PIV technique has been converted to the micro scale by the use of stereo-microscopy. In this work a different technique is investigated, which uses conventional, two-component micro-PIV to generate velocity data on a number of planes. The in-plane velocity gradients are then calculated, which can be used in the continuity equation to produce the out-of-plane velocity gradients. These, together with the no-penetration boundary condition, can then be used to calculate the out-of-plane velocities. An algorithm is presented that is capable of handling up to one invalid vector per column of data by using a combination of first order and second order projections of the velocity. The advantage of the continuity based technique is that it uses the existing two-component micro-PIV technology, which at present is in a more advanced stage of development then stereo-microscopy based micro-PIV. The technique is investigated using a flow similar to one used previously to assess stereoscopic micro-PIV (Meas Sci Technol 17:2175–2185, 2006). This allows a comparison of the performance of the two techniques. The results show that the continuity based data agrees well with an independent computational fluid dynamics solution and has a smaller experimental uncertainty than the stereoscopic technique at a better spatial resolution. There are, however, potential limitations to the continuity based technique. These include the two-dimensionality of the data, which is limited to the planes on which the original images were taken, and the dependence of the technique on the data close to surfaces, where the experimental errors are often greatest. Stereoscopic micro-PIV does not have these limitations so, whilst at present it appears that continuity based techniques may be more accurate, there is sufficient potential for stereoscopic techniques to justify their further development.     

Constitutive Model for the Thermo-viscoplastic Behavior of Hexagonal Close-Packed Metals with Application to Ti-6A1-4V Alloy

In this paper, a new physically based constitutive model is developed for hexagonal close-packed metals, especially the Ti-6 A1-4V alloy, subjected to high strain rate and different temperatures based on the microscopic mechanism of plastic deformation and the theory of thermally activated dislocation motion. A global analysis of constitutive parameters based on the Latin Hypercube Sampling method and the Spearman's rank correlation method is adopted in order to improve the identification efficiency of parameters. Then, an optimal solution of constitutive parameters as a whole is obtained by using a global genetic algorithm composed of an improved niche genetic algorithm, a global peak determination strategy and the local accurate search techniques. It is concluded that the proposed constitutive modal can accurately describe the Ti-6 Al-4V alloy's dynamic behavior because the prediction results of the model are in good agreement with the experimental data.     

Consistency between independence theorems and generalized self-consistent method

Recently Zheng & Hwang established a series of independence theorems concerning with planar effective elastic properties. It is manifested that the estimation of the effective elastic properties of microcracked solids through the generalized self-consistent method (GSCM) contradicts with these independence theorems. In this paper it is shown that such contradiction is actually caused by the approximate algorithm adopted, while the exact solution of GSCM is consistent with these rigorously established independence theorems. Since only an approximate algorithm in GCSM is available in dealing with problems involving non-circular inclusions or holes, an intrinsic GSCM is proposed, which can be performed based on an approximate algorithm and the corresponding estimations are consistent with the independence theorems.     

Assessment of adaptive and heuristic time stepping for variably saturated flow

The performance of improved initial estimates and ‘heuristic’ and ‘adaptive’ techniques for time step control in the iterative solution of Richards equation is evaluated. The so‐called heuristic technique uses the convergence behaviour of the iterative scheme to estimate the next time step whereas the adaptive technique regulates the time step on the basis of an approximation of the local time truncation error. The sample problems used to assess these various schemes are characterized by nonuniform (in time) boundary conditions, sharp gradients in the infiltration fronts, and discontinuous derivatives in the soil hydraulic properties. It is found that higher order initial solution estimates improve the convergence of the iterative scheme for both the heuristic and adaptive techniques, with greater overall performance gains for the heuristic scheme, as could be expected. It is also found that the heuristic technique outperforms the adaptive method under strongly nonlinear conditions. Previously reported observations suggesting that adaptive techniques perform best when accuracy requirements on the numerical solution are very stringent are confirmed. Overall both heuristic and adaptive techniques have their limitations, and a more general or mixed time stepping strategy combining truncation error and convergence criteria is recommended for complex problems. Copyright © 2006 John Wiley & Sons, Ltd.     

Analysis of entry flow to determine elongation flow properties revisited

The minimisation technique proposed by Binding (J. Non-Newtonian Fluid Mech., 27 (1988) 173) was used in our Generalised Engineering Bernoulli Equation framework to relate the entry pressure and stress power. We arrived at a final result similar to Binding's using assumed kinematics. Through subsequent assumptions to the kinematics we finally arrive at a result exactly equivalent to Cogswell's technique (Trans. Soc. Rheol., 16 (1972) 383). Thus, these two techniques are related in this general framework. The techniques were used to predict elongation flow properties of a polymer melt and polymer solution. The results for the polymer melt clearly show Cogswell's technique is adequate at high elongation rates. All these techniques require minimisation of the stress power with respect to the flow volume and discussion is given as to the validity of this minimisation technique. In addition, the approximate variational technique we propose gives clears limits as to when a technique, such as Cogswell's, can be applied.     

Efficient and stable spectral element methods for predicting the flow of an XPP fluid past a cylinder

Stable and accurate spectral element methods for predicting the flow of branched polymer melts past a confined cylinder are presented. The fluid is modelled using a modification of the pom–pom model known as the extended pom–pom (XPP) model. Steady and transient flows are considered in this paper. The operator integration factor splitting technique is used to discretize the governing equations in time, while the spectral element method is used in space. An iterative solution algorithm that decouples the computation of velocity and pressure from that of stress is used to solve the discrete equations. Appropriate preconditioners are developed for the efficient solution of these problems. Local upwinding factors are used to stabilize the computations. Numerical results are presented demonstrating the performance of the algorithm and the predictions of the model. The influence of the model parameters on the solution is described and, in particular, the dependence of the drag on the cylinder as function of the Weissenberg number.     

Semi-implicit finite difference methods for three-dimensional shallow water flow

A semi-implicit finite difference method for the numerical solution of three-dimensional shallow water flows is presented and discussed. The governing equations are the primitive three-dimensional turbulent mean flow equations where the pressure distribution in the vertical has been assumed to be hydrostatic. In the method of solution a minimal degree of implicitness has been adopted in such a fashion that the resulting algorithm is stable and gives a maximal computational efficiency at a minimal computational cost. At each time step the numerical method requires the solution of one large linear system which can be formally decomposed into a set of small three-diagonal systems coupled with one five-diagonal system. All these linear systems are symmetric and positive definite. Thus the existence and uniquencess of the numerical solution are assured. When only one vertical layer is specified, this method reduces as a special case to a semi-implicit scheme for solving the corresponding two-dimensional shallow water equations. The resulting two- and three-dimensional algorithm has been shown to be fast, accurate and mass-conservative and can also be applied to simulate flooding and drying of tidal mud-flats in conjunction with three-dimensional flows. Furthermore, the resulting algorithm is fully vectorizable for an efficient implementation on modern vector computers.