频率约束板结构拓扑优化

本文利用ICM（独立、连续、映射）方法建立了频率约束下平板结构重量最轻的拓扑优化模型。采用指数函数作为单元重量、单元质量及单元刚度的过滤函数。通过瑞利商对刚度过滤函数倒变量的泰勒一阶展式,将频率约束近似显式化。利用对偶理论将含有大量设计变量的约束优化模型转化为易于求解的少设计变量拟无约束优化模型,通过序列二次规划将转化模型进行求解,提高了求解的效率。本文选择MSC.Patran&Nastran软件及PCL二次开发语言构架了平板结构频率约束拓扑优化问题的软件。数值算例表明：本文的方法具有迭代稳定性和收敛高效性。     

一种基于Neumann级数的区间有限元方法

实际工程问题中通常存在大量的不确定参数, 区间有限元方法是一种结合有限元数值计算工具对结构进行不确定性分析的区间方法. 区间有限元的目的是获得在含有区间不确定性参数条件下的结构响应上下边界, 其关键问题在于区间平衡方程组的求解, 而这属于一类往往很难求解的NP-hard问题. 本文归纳了一类工程实际中常见的结构不确定性问题, 即可线性分解式区间有限元问题, 并针对此提出一种基于Neumann级数的区间有限元方法. 在区间有限元分析中, 当区间不确定参数表示为一组独立区间变量线性叠加时, 若结构的刚度矩阵也可表示为这些独立区间变量的线性叠加形式, 则称此类区间有限元问题为可线性分解式区间有限元问题. 对于此类问题, 采用Neumann级数对其刚度矩阵的逆矩阵进行表示, 可获得结构响应关于区间变量的显式表达式, 从而可高效求解结构响应的上下边界. 最后通过两个算例验证了本文所提方法的有效性.      

AN INTERVAL FINITE ELEMENT METHOD BASED ON THE NEUMANN SERIES EXPANSION 1)

实际工程问题中通常存在大量的不确定参数, 区间有限元方法是一种结合有限元数值计算工具对结构进行不确定性分析的区间方法. 区间有限元的目的是获得在含有区间不确定性参数条件下的结构响应上下边界, 其关键问题在于区间平衡方程组的求解, 而这属于一类往往很难求解的NP-hard问题. 本文归纳了一类工程实际中常见的结构不确定性问题, 即可线性分解式区间有限元问题, 并针对此提出一种基于Neumann级数的区间有限元方法. 在区间有限元分析中, 当区间不确定参数表示为一组独立区间变量线性叠加时, 若结构的刚度矩阵也可表示为这些独立区间变量的线性叠加形式, 则称此类区间有限元问题为可线性分解式区间有限元问题. 对于此类问题, 采用Neumann级数对其刚度矩阵的逆矩阵进行表示, 可获得结构响应关于区间变量的显式表达式, 从而可高效求解结构响应的上下边界. 最后通过两个算例验证了本文所提方法的有效性.     

开口薄壁杆件结构稳定分析的精确单元和两步求解算法

从控制微分方程的通解出发,构造受偏心压力作用开口薄壁杆件的精确形函数,建立用于开口薄壁杆件结构稳定性分析的精确有限元,得到了单元刚度矩阵和几何刚度矩阵的显式表达,提出了计算给定区间内各阶临界荷载以及相应失稳模态的两步计算方法。计算结果表明,与常规单元相比,采用精确单元无需进行网格细分就可以获得精确的数值结果,结合本文的两步求解算法,可以准确获得给定区间内全部临界荷载和失稳模态。     

A novel beam finite element with singularities for the dynamic analysis of discontinuous frames

In this work, a model of the stepped Timoshenko beam in presence of deflection and rotation discontinuities along the span is presented. The proposed model relies on the adoption of Heaviside’s and Dirac’s delta distributions to model abrupt and concentrated, both flexural and shear, stiffness discontinuities of the beam that lead to exact closed-form solutions of the elastic response in presence of static loads. Based on the latter solutions, a novel beam element for the analysis of frame structures with an arbitrary distribution of singularities is here proposed. In particular, the presented closed-form solutions are exploited to formulate the displacement shape functions of the beam element and the relevant explicit form of the stiffness matrix. The proposed beam element is adopted for a finite element discretization of discontinuous framed structures. In particular, by means of the introduction of a mass matrix consistent with the adopted shape functions, the presented model allows also the dynamic analysis of framed structures in presence of deflection and rotation discontinuities and abrupt variations of the cross-section. The presented formulation can also be easily employed to conduct a dynamic analysis of damaged frame structures in which the distributed and concentrated damage distributions are modelled by means of equivalent discontinuities. As an example, a simple portal frame, under different damage scenarios, has been analysed and the results in terms of frequency and vibration modes have been compared with exact results to show the accuracy of the presented discontinuous beam element.     

桁架拓扑优化的多点逼近遗传算法

提出一种基于多点逼近函数和遗传算法的桁架拓扑优化方法。该方法建立了包含连续尺寸和离散拓扑两类变量的优化模型，并通过构造多点逼近函数建立了结构优化问题的第一级序列显式近似，然后采用分层优化方法：在外层对拓扑变量采用遗传算法进行优化；在内层对尺寸变量通过可由对偶法求解的第二级序列近似问题进行优化。几个经典的桁架拓扑优化算例表明该方法能以较少的结构分析次数获得比较理想的概率意义上的最优解。     

Plate/shell topological optimization subjected to linear buckling constraints by adopting composite exponential filtering function

In this paper, a model of topology optimization with linear buckling constraints is established based on an independent and continuous mapping method to minimize the plate/shell structure weight. A composite exponential function (CEF) is selected as filtering functions for element weight, the element stiffness matrix and the element geomet-ric stiffness matrix, which recognize the design variables, and to implement the changing process of design variables from“discrete”to“continuous”and back to“discrete”. The buck-ling constraints are approximated as explicit formulations based on the Taylor expansion and the filtering function. The optimization model is transformed to dual programming and solved by the dual sequence quadratic programming algo-rithm. Finally, three numerical examples with power function and CEF as filter function are analyzed and discussed to demonstrate the feasibility and efficiency of the proposed method.     

基于初弯曲单元的某弦支穹顶非线性稳定承载力分析

在单元的随动坐标系下,建立了初始弯曲对杆单元轴向刚度影响的计算公式,导出了考虑初始弯曲的非线性杆单元刚度矩阵;从经典的梁-柱理论出发,给出了考虑初始弯曲的非线性空间梁单元切线刚度矩阵推导过程。将建立的初始弯曲单元应用于弦支穹顶结构算例的非线性稳定承载力分析,研究了杆件初始弯曲对结构整体非线性稳定性能的影响,结果表明,杆件初始弯曲的存在会降低结构的整体刚度和极限承载力,而在不同的初始预应力状态下,极限荷载随初始弯曲增加都基本呈线性下降,并且其关系曲线基本平行。本文建立的基于初弯曲单元的非线性分析方法,可推广应用于各种类型空间网格结构的非线性分析,为结构设计方案的选择提供更为科学的依据。     

The analytical solutions based on the concept of finite element methods

On the basis of the concept of finite element methods, the rigorous analytical solutions of structural response in terms of the design variables are researched in this paper. The spatial trusses are taken as an example for the solution of the analytical expressions of the explicit displacements which are proved mathematically; then some conclusions are reached that are useful to structural sensitivity analysis and optimization. In the third part of the paper, a generalized geometric programming method is sugguested for the optimal model with the explicit displacement. Finally, the analytical solutions of the displacements of three trusses are given as examples.     

四节点二十四自由度平板壳单元几何刚度矩阵显式解析式的推演算法研究

几何刚度矩阵的推演是结构几何非线性有限元分析的重点和难点之一。推导几何刚度矩阵显式解析表达式成为简化非线性有限元列式,提高分析效率的关键。本文在协同转动法框架下,基于刚体运动法则对四节点二十四自由度的平板壳单元几何刚度矩阵显式解析式进行了推导和讨论;分析了悬臂梁大转动、不同壁厚条件下简支圆柱形屋顶空间大变位两个经典算例。研究结果表明：（1）几何刚度矩阵的显式计算公式不仅为板壳结构几何非线性列式提供了方便而且具有良好的精度;（2）推导的几何刚度矩阵适用于各类型四边形二十四自由度平板壳单元模型;（3）与数值积分相比,采用解析形式的几何刚度矩阵可以显著提高非线性响应计算效率。     

Component-wise analysis of laminated anisotropic composites

This paper proposes a one-dimensional (1D) refined formulation for the analysis of laminated composites which can model single fibers and related matrices, layers and multilayers. Models built by means of an arbitrary combination of these four components lead to a component-wise analysis. Different scales can be used in different portions of the structure and this leads to a global–local approach. In this work, computational models were developed in the framework of finite element approximations. The 1D FE formulation used has hierarchical features, that is, 3D stress/strain fields can be detected by increasing the order of the 1D model used. The Carrera Unified Formulation (CUF) was exploited to obtain advanced displacement-based theories where the order of the unknown variables over the cross-section is a free parameter of the formulation. Taylor- and Lagrange-type polynomials were used to interpolate the displacement field over the element cross-section. Lagrange polynomials permitted the use of only pure displacements as unknown variables. The related finite element led straightforwardly to the assembly of the stiffness matrices at the structural element interfaces (matrix-to-fiber, matrix-to-layer, layer-to-layer etc). Preliminary assessments with solid model results are proposed in this paper; various numerical examples were carried out on cross-ply symmetrical fiber-reinforced laminates [0/90/0] and a more complex composite C-shaped model. The examples show that the proposed models can analyze laminated structures by combining fibers, matrices, layers and multilayers and by referring to a unique structural finite element formulation.     

改进的小损伤结构动力学有限元建模方法

在遗传算法或神经网络方法识别结构损伤位置和程度时,都是基于少量的在线测量的损伤结构振动响应数据和大量的模型仿真数据来实现的,因而建立高效和精确的损伤结构动力学有限元模型,以便仿真获得损伤结构的大量动力学响应数据是十分重要的基础前提工作。本文针对ANSYS结构分析软件在建立结构小损伤有限元动力学模型存在两个关键问题,结构损伤处直接网格划分的计算结果误差和网格节省问题,以结构损伤振动检测的实际需要为出发点,提出了建立含小损伤结构的ANSYS动力学建模技术,研究了结构局部小损伤及其位置与所在处单元刚度矩阵变化的数量关系。     

Topology optimization design of continuum structures under stress and displacement constraints

ＩｎｔｒｏｄｕｃｔｉｏｎＴｏｐｏｌｏｇｙｏｐｔｉｍｉｚａｔｉｏｎｏｆｃｏｎｔｉｎｕｕｍｓｔｒｕｃｔｕｒｅｓｄｉｄｎｏｔｄｅｖｅｌｏｐｒａｐｉｄｌｙｕｎｔｉｌｒｅｃｅｎｔｔｅｎｙｅａｒｓｏｗｉｎｇｔｏｔｈｅｓｐｅｃｉａｌｄｉｆｆｉｃｕｌｔｉｅｓｉｎｖｏ...     

压电桁架作动器/传感器优化配置算法研究

针对自适应压电桁架结构振动控制,建立了作动器/传感器优化配置数学模型,并提出一种优化配置的新方法。为了减少结构分析次数,该方法将近似概念、对偶法和遗传算法相结合,首先采用多点近似技术建立原问题的序列近似问题,再对近似问题中的作动器/传感器位置离散变量和控制增益连续变量采用遗传算法和对偶方法分别寻优的分层优化策略。为了提高近似问题对原问题的逼近程度,本文提出一种适于离散变量结构优化的分段多点近似函数。算例表明本文方法能够以很少的结构分析次数得到最优解。     

An hysteretic quadrilateral plane stress element

In this work, a new plane stress element is proposed for the nonlinear static and dynamic analysis of plane stress/plane strain problems. The four node quadrilateral element formulation for the elastic case is extended by introducing a novel hysteretic constitutive relation, based on the Bouc–Wen model of hysteresis. The hysteretic model introduced is directly derived from the governing equations of classical plasticity based on the flow rule and specific hardening law. The stiffness matrix of the element is formulated using the principle of virtual displacements, where the elastic stress–strain relation is substituted by the hysteretic relation proposed. The derived stiffness matrix is expressed as a smooth function of the internal stress field both in the elastic and inelastic regime. The efficiency of the proposed element in the simulation of the cyclic behavior in plane structures is presented through illustrative examples.     

多位移等式约束拓扑优化及在光电精密结构设计中的应用

为了满足光电精密跟踪设备中光学系统对支撑结构变形位移相等的设计要求,基于变密度法,以刚度极大为目标,同时以体积约束和位移等式约束作为约束条件,构建结构拓扑优化模型。位移等式约束通过增广拉格朗日乘子法引入原目标函数,在拉格朗日乘子的求解中,采用考虑具有真实物理意义的近似替代法而非传统的纯数学迭代逼近方法。在利用伴随方法得到增广目标函数敏度基础上,采用MMA优化算法,在满足体积约束的同时进行迭代优化得到新结构。算例验证结果表明,本文方法能够有效解决具有多个位移等式约束的刚度极大结构轻量化设计问题。     

适用于壳体h型自适应有限元分析的一组新单元

推导出一组适用于h型自适应分析的四边形蜕化壳元。对于大多数壳体结构,壳单元的刚度矩阵可分为薄膜、弯曲和剪切三部分。对薄膜部分本文采用杂交应力元方法进行设计,独立假设薄膜应力场以改善其精度;弯曲部分的刚度矩阵则依然由基于位移的应变来获得;而剪切部分则采用假设自然应变的方法来获得能克服薄壳下剪切自锁的新剪应变并用于计算此部...     

框架结构屈曲的精确有限元求解

基于屈曲微分控制方程的一般解,构造了Euler梁在轴力作用下的精确形函数,建立了用于框架结构屈曲分析的精确有限单元,得到了单元刚度矩阵和几何刚度矩阵的显式表达,并提出了基于常规特征值计算的迭代算法以确定屈曲载荷及相应失稳模态的精确解. 研究表明, 对于线性稳定性分析而言,常规框架有限单元可视为精确有限单元的一种近似. 若采用精确单元,无需进行网格细分就可以获得精确的屈曲载荷和失稳模态. 数值算例证明了新单元和算法的效率和精度.      

多约束膜结构截面优化及在MSC/Nastran上的程序实现

将准则法和数学规划法相结合，借助满应力准则将应力约束转化为动态尺寸约束，利用单位虚载荷法将位移约束转化为设计变量的显式表达式建立优化模型，然后用数学规划法求解；采用无量纲设计变量实现设计变量连接，对膜结构的厚度进行优化设计；根据对偶理论，应用对偶规划精确映射原问题，再按泰勒展式建立对偶问题的二阶近似。为了提高优化效率，采用射线步调整结构性态，运用粗选有效约束技术筛选约束，并采用主、被动变量循环确保收敛稳定。以MSC／Nastran软件作为结构分析的求解器，以MSC／Patran软件作为开发平台，完成了膜结构截面优化程序。对膜结构的单变位、多变位的结构优化问题进行了优化计算，并与MSC／Nastran优化模块的计算结果进行比较。算例结果表明程序的可靠性、高效性和稳定性以及理论算法的优越性。     

Functional and generalized separable solutions to unsteady Navier–Stokes equations

The paper studies unsteady Navier–Stokes equations with two space variables. It shows that the non-linear fourth-order equation for the stream function with three independent variables admits functional separable solutions described by a system of three partial differential equations with two independent variables. The system is found to have a number of exact solutions, which generate new classes of exact solutions to the Navier–Stokes equations. All these solutions involve two or more arbitrary functions of a single argument as well as a few free parameters. Many of the solutions are expressed in terms of elementary functions, provided that the arbitrary functions are also elementary; such solutions, having relatively simple form and presenting significant arbitrariness, can be especially useful for solving certain model problems and testing numerical and approximate analytical hydrodynamic methods. The paper uses the obtained results to describe some model unsteady flows of viscous incompressible fluids, including flows through a strip with permeable walls, flows through a strip with extrusion at the boundaries, flows onto a shrinking plane, and others. Some blow-up modes, which correspond to singular solutions, are discussed.