遗传算法求解可行域分离的结构优化问题

应用遗传算法求解了两类可行域分离的结构优化问题：局部屈曲约束的桁架拓扑优化问题和动力响应约束优化问题．对第一类问题，提出了新的数学表达式，适合于遗传算法求解．采用了改进的适应度函数及约束处理方法、约束凝聚选择、交叉操作改进和竞争最优保留，提高了遗传算法的效率和可靠性．算例说明，该方法能够克服可行域分离给传统优化算法带来的困难，有效地在多连通可行域中搜索全局最优解．     

单侧接触约束下桁架拓扑优化的若干结果

对考虑应力和无摩擦单侧接触位移约束的桁架拓扑优化问题进行了研究,采用解析方法,通过对典型算例不同初始接触间隙下的拓扑优化解的讨论,发现这一优化问题中存在解不唯一现象,研究也表明,接触状态对结构拓扑有一定影响,不同初始接触间隙可能导致不同的拓扑结构,对利用数值方法该问题进行了研究,为克服结构拓扑优化中奇异最优解问题,采用ε松弛方法建立桁架拓扑优化模型列式,并利用双层优化方法求解本拓扑优化问题,第一层,通过求解与原接触问题等价的二次规划问题来进行结构接触分析;第二层,利用第一层得到的位移及应力,采用ε-松弛方法进行结构拓扑优化。     

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。     

离散变量结构拓扑优化设计研究

研究了离散变量结构拓扑优化设计的若干问题，讨论了离散型优化模型的合理性，提出截面设计变量的离散程度和全局约束影响最优拓扑，是优化中不可忽视的因素，文中还提出了一种解离散变量桁架，刚架结构拓扑优化的启发式算法。     

应力和位移约束下桁架拓扑优化的有无复合体方法

按照独立连续拓扑变量的思想,将有无复合体模型由构造应力约束问题发展为应力与位移约束问题,用于桁架结构拓扑优化．基于该文提出的位移约束与拓扑变量关系的近似显式,建立了具有应力、位移约束下桁架结构拓扑优化的有无复合体模型,用序列二次规划算法求解．数值结果令人满意．这一工作表明独立连续拓扑变量的提出对于结构拓扑优化的研究是有价值的。     

改进的遗传算法求解桁架的拓扑优化

提出了一种改进的遗传算法,用于优化具有离散尺寸、连续形状和0-1拓扑变量的桁架问题。考虑到离散和连续变量的本质,文中提出了混合编码方法,其中包括二进制和实数编码,整数和实数编码。本文采用了凝聚选择法-基于约束和适应度值双重标准,完全适应约束问题的本质。在优化过程中,初始种群和算子具有不确定性,因此有必要检验结构拓扑的合理性。为了增强算法的可靠性,采用了改进的重新开始算子,引入新基因并且探索新空间。求解了典型的算例,证明改进的遗传算法是可行且有效的。     

多工况多约束下离散变量桁架结构的拓扑优化设计

提出了一个多工况下受应力、位移约束的离散变量桁架结构的拓扑优化方法，给出了结构拓扑形式变更时的约束处理方法及杆件删除策略，使基结构设计空间的维数不断降低，达到最优拓扑，避免了奇异解的出现，在算例中指出了截面离散集和位移约束对最优拓扑的影响；算例给出了满意的拓扑优化解。     

演化算法求解桁架多目标拓扑优化的收敛速度研究

演化算法能够同时满足结构拓扑优化的前沿领域对全局优化、黑箱函数优化、组合优化和多目标优化的需求,但采用此类算法的可行性与必要性由其收敛性与计算效率决定。本文以应力约束桁架多目标拓扑优化问题为求解对象,致力于揭示在收敛性与计算效率两方面具有竞争力的算法。首先提出评估演化算法求解拓扑优化问题收敛性与计算效率的通用方法,采用穷举法严格推导了典型桁架多目标拓扑优化问题的全局最优解,并采用超体积指标定义了多层次收敛性能准则。最后通过比较研究得到不同收敛性需求下具有最快收敛速度的演化算法,并揭示了具有竞争力的算法机制。本研究为演化算法求解多目标拓扑优化问题的收敛速度奠定了理论基础,同时为高效求解实际工程拓扑优化问题提供算法支持。     

平面膜结构拓扑优化的有无复合体方法

将作者对桁架在应力约束下结构拓扑优化的有无复合体模型发展到平面膜结构在应力、位移约束下结构拓扑优化的建模与求解。同时提出了该模型的有效解法,获得了令人满意的数值结果。本文工作表明独立连续拓扑变量的提出对于结构拓扑优化的研究是有意义的。     

基于和声搜索遗传算法的桁架结构拓扑优化

为改善传统设计理念和遗传算法优化不足,促进桁架结构离散变量拓扑优化发展与创新,将遗传算法与和声搜索算法混合,同时对遗传交叉和变异分3种情况进行自适应改进,建立了用于桁架结构拓扑优化的新型混合遗传算法——和声搜索遗传算法,利用该方法分别对平面桁架和空间桁架结构进行拓扑优化分析,并与改进遗传算法、拟满应力遗传算法、相对差商法、复合形遗传算法和改进蚁群算法比较,证明了此方法是有效、可行的。     

一种基于直接计算高阶奇异积分的断裂力学双边界积分方程分析法

相对于有限元法,边界单元法在求解断裂问题上有着独特的优势,现有的边界单元法中主要有子区域法和双边界积分方程法.采用一种改进的双边界积分方程法求解二维、三维断裂问题的应力强度因子,对非裂纹边界采用传统的位移边界积分方程,只需对裂纹面中的一面采用面力边界积分方程,并以裂纹间断位移为未知量直接用于计算应力强度因子.采用一种高阶奇异积分的直接法计算面力边界积分方程中的超强奇异积分;对于裂纹尖端单元,提供了三种不同形式的间断位移插值函数,采用两点公式计算应力强度因子.给出了多个具体的算例,与现存的精确解或参考解对比,可得到高精度的计算结果.      

汶川地震诱发文家沟巨型滑坡 碎屑流基本特征及成因机制初步分析


2008年5月12日,汶川M80地震在四川省绵竹市清平乡文家沟内诱发一巨型滑坡。通过现场调查得知,滑坡前后缘高差455m,厚度20～30m,滑面为基岩层面,初始方量2750×107m3。滑体在运动中转化为碎屑流。滑坡-碎屑流总的水平运动距离为4022m,垂直运动距离为1443m,遗留的堆积物体积达5×107m3。滑坡距映秀—北川断裂仅36km,位于其下盘,地震烈度达XI度。滑坡导致文家沟中48人遇害,并形成一条完整的地震次生地质灾害链。初步分析表明滑坡启动速度快,滑坡向碎屑流转化过程明显、地点明确。碎屑流运动过程复杂,伴有强烈的“气垫效应”和“前缘气浪冲击效应”。作者认为,文家沟滑坡的高启动速度是长持时强烈地震动作用的结果,与山体的猛烈碰撞是导致滑体解体并转化为碎屑流的原因。     

Preface: Orbits around asteroid

One of the core issues in modern celestial mechanics is the orbital dynamics in the near-regime gravitational field of as- teroids, which provides deep insights into the mathematical nature of a class of nonlinear systems, and plays as a critical basis for in situ explorations of different science goals. Lots of efforts have been made to reveal the characteristics of orbital motion in the vicinity of asteroids, and to improve the skills of asteroid research in methodology.     

Effet de l'anisotropie élastique cristalline sur la distribution des facteurs de Schmid à la surface des polycristauxInfluence of crystalline elasticity anisotropy on Schmid factor distribution at the free surface of polycrystals

Experimental studies of the plasticity mechanisms of polycrystals are usually based on the Schmid factor distribution supposing crystalline elasticity isotropy. A numerical evaluation of the effect of crystalline elasticity anisotropy on the apparent Schmid factor distribution at the free surface of polycrystals is presented. Cubic elasticity is considered. Order II stresses (averaged on all grains with the same crystallographic orientation) as well as variations between averages computed on grains with the same crystallographic orientation but with different neighbour grains are computed. The Finite Element Method is used. Commonly studied metals presenting an increasing anisotropy degree are considered (aluminium, nickel, austenite, copper). Concerning order II stresses in strongly anisotropic metals, the apparent Schmid factor distribution is drifted towards small Schmid factor values (the maximum Schmid factor is equal to 0.43 instead of 0.5) and the slip activation order between characteristic orientations of the crystallographic standard triangle is modified. The computed square deviations of the stresses averaged on grains with the same crystallographic orientation but with different neighbour grains are a bit higher than the second order ones (inter-orientation scatter). Our numerical evaluations agree quantitatively with several observations and measures of the literature concerning stress and strain distribution in copper and austenite polycrystals submitted to low amplitude loadings. Hopefully, the given apparent Schmid factor distributions could help to better understand the observations of the plasticity mechanisms taking place at the free surface of polycrystals. To cite this article: M. Sauzay, C. R. Mecanique 334 (2006).     

Nonlinear behavior of matrix-inclusion composites under high confining pressure: application to concrete and mortar

This paper is devoted to a micromechanics-based simulation of the response of concrete to hydrostatic and oedometric compressions. Concrete is described as a composite made up of a cement matrix in which rigid inclusions are embedded. The focus is put on the role of the interface between matrix and inclusion which represent the interfacial transition zone (ITZ). A plastic behavior is considered for both the matrix and the interfaces. The effective response of the composite is derived from the modified secant method adapted to the situation of imperfect interfaces. To cite this article: T.H. Le et al., C. R. Mecanique 336 (2008).     

Dynamics and Breakup of Viscoelastic Liquids (A Review)

The phenomena of hydrodynamic breakup of liquid jets, drops, films, bridges, and filaments are reviewed for liquids with viscoelastic properties. The reasons for breakup are capillary instabilities, collisions with rigid obstacles, and other forms of dynamic action. The relationship between the properties of the liquids and the features of the breakup process is discussed.     

Flow stability analysis and excitation using pulsating jets

Classical flow stability applied to transition from laminar to turbulent flow may also describe the behavior of vorticity fluctuations created by a pulsating jet placed along a solid boundary. A numerical laminar flow experiment involving a pulsating jet placed along the surface of a duct with flow separation downstream, resulted in eliminating most part of the separated flow region. Applying the same approach to a turbulent flow, it was possible to develop a turbulent stability flow formulation and apply successfully turbulent pulsating jet flow separation control. To cite this article: D. Skamnakis, K. Papailiou, C. R. Mecanique 333 (2005).