Using discrete optimization algorithms to find minimum energy configurations of slender cantilever beams with non-convex energy functions

This paper deals with a new solution technique for approximately solving certain variational problems in elasticity by using discrete optimization techniques that were originally used in information theory. This allows us to easily and approximately solve large deformation buckling problems for slender cantilever beams (including post-buckling behavior) as well as problems where the strain energy function is non-convex.The core idea is to quantize or discretize the variables describing the possible configurations of the body. This, when combined with the fact that the variational problem has an inherent Markov structure allows us to use computationally efficient search techniques based on dynamic programming (equivalent to finding the shortest path in a weighted directed graph) to find optimal solutions within the quantized state space. The results can be used in two ways: (1) directly as a fast approximate solution to the variational problem (2) As a means for finding very good (nearly minimum energy) initial configurations for application of conventional minimization techniques, which might otherwise fail because of a poor starting configurations which are far from the global minimum. We demonstrate both these uses in the paper.     

New methods to find solutions and analyze stability of equilibrium of nonholonomic mechanical systems

A large proportion of constrained mechanical systems result in nonlinear ordinary differential equations, for which it is quite difficult to find analytical solutions. The initial motions method proposed by Whittaker is effective to deal with such problems for various constrained mechanical systems, including the nonholonomic systems discussed in the first part of this paper, where in addition to differential equations of motion, nonholonomic constraints apply. The final equations of motion for these systems are obtained in the form of corresponding power series. Also, an alternative, direct method to determine the initial values of higher-order derivatives \({\ddot{q}}_0 ,{{\dddot{q}{} }}_{\!0} ,\ldots \) is proposed, being different from that of Whittaker. The second part of this work analyzes the stability of equilibrium of less complex, nonholonomic mechanical systems represented by gradient systems. We discuss the stability of equilibrium of such systems based on the properties of the gradient system. The advantage of this novel method is its avoidance of the difficulty of directly establishing Lyapunov functions aimed at such unsteady nonlinear systems. Finally, these theoretical considerations are illustrated through four examples.     

基于弧长法的受压球壳稳定性分析

对受压球壳进行特征值屈曲分析,得到了前6阶屈曲模态及线性屈曲临界载荷;采用弧长法进行非线性有限元分析,对理想球壳施加初始扰动,通过2次扰动值折半的方法求得引起结构屈曲的最小扰动值,追踪到了屈曲分支点和全过程载荷-位移路线。基于前6阶屈曲模态位移,在受压球壳中分别引入2.5mm和1mm两种缺陷值,分析缺陷对球壳屈曲特性的影响。结果表明:取壳厚的0.5%即0.05mm时,得最小扰动值,近似模型与完善结构极值载荷的差值为0.93%;球壳是缺陷敏感性结构,缺陷的幅值和分布都对其极限载荷有影响,缺陷幅值与厚度比为0.1时,缺陷球壳承载力相对理想结构下降了约11%,缺陷幅值与厚度比为0.25时,承载力相对下降了约30%,说明提高球壳稳定性需要提高球壳加工精度。     

Identification of Parameters of Short-Time Creep of Plexiglas By Means of Studying Decaying Flexural Vibrations of Test Samples

A technique for identification of hereditary properties under conditions of short-time creep of Plexiglas is developed. The technique is based on experimental determination of shifting of the center of decaying flexural vibrations of vertically aligned test samples after their preliminary maintaining in a static bent state. Mathematically, the technique is based on using the finite element method and integral equations of the theory of hereditary viscoelasticity with the Koltunov–Rzhanitsyn hereditary kernel. An object function is constructed for identification of rheological parameters of this kernel. The minimum of this function is found by the direct search method, which does not require the function gradient to be known. The hereditary kernel of Plexiglas averaged on the basis of data for several test samples is obtained as a function of time.     

Extremum principles for certain hyperelastic materials

A hyperelastic material is here said to be of class H_m if the elastic potential is a homogeneous function of order m + 1 in the components of the Lagrangian displacement gradient. It is shown that a single solution to a boundary value problem generates an infinite family of solutions to a family of related boundary value problems. Assuming that a solution to a boundary value problem exists, it is shown that it is unique provided that the material is stable in the sense of Hill in a deleted neighbourhood of the stress-free state. A minimum theorem concerning the strain energy and the virtual work of the prescribed forces is established for the equilibrium configurations, and a maximum theorem concerning the virtual work of the prescribed surface displacements and the complementary stress energy is established for compatible stress fields. As an application, upper and lower bounds are found for the torsional stiffness of a cylindrical bar of square cross section under infinitesimal deformation.     

以优化方法求解布料无摩擦接触变形问题

利用作者所提出的布料质点模型,用罚函数法解除布料和其它物体之间的接触约束,把接触约束条件下的极值问题转化为无约束极值问题,用F-R共轭梯度法求解,并用三次插值方法进行一维搜索．计算了有缝摺和裁剪开口的布料与球形物体的接触变形问题,得到的结果符合人们的日常实际观察．     

On the use of the complementary energy in the solution of buckling problems

A systematic derivation of the expression for the complementary energy in elastic buckling problems is presented. Compatibility is identified with variation with respect to the stress components, and the resulting eigenvalue problem is shown to be equivalent to, and sometimes more convenient than, the corresponding formulation in terms of the potential energy. Similarly, approximate techniques may lead to better as well as simpler estimates, whose upper bound property can, however, be assured only through appropriate safeguards.The method is applied in some detail to buckling of columns of arbitrary boundary conditions and axial force distribution. Another example is the problem of lateral beam buckling, with the effect of warping restraint included. In both cases (and presumably in many others) the complementary energy formulation is of lower order than the conventional potential energy formulation, and it is clear that the same simplification should also apply to finite elements or other discrete formats. The method is restricted to the (technically significant) case of a linear prebuckling state.     

计及表面能的应变梯度理论本构参数识别

本文提出利用静态位移信息对一种计及表面能的应变梯度理论本构参数进行识别的求解策略.基于Vardoulakis和Sulem的计及表面能的简单线性应变梯度理论,文献[13]给出了伯努力-欧拉梁弯曲问题的正演解析模型,本文将其反演归结为两个带有不等式约束的非线性规划问题.在此基础上,采用黄金分割一维搜索方法进行求解,给出了数值验证,讨论了信息误差对反演结果的影响.结果显示,这种方法可以用来对应变梯度理论本构参数进行识别,即使在体积和表面能常数非常小的情况下,仍然能够得到满意的结果.     

Lateral instability of rectangular plates

This paper investigates the problems of lateral buckling of rectangular plates. In the text we discuss the minimum critical load of the lateral buckling occurring on under a concentrated force, uniformly distributed load and the concentrated couples, respectively. The energy method is used in this article.     

Three-dimensional equilibria of nonlinear pre-curved beams using an intrinsic formulation and shooting

This article describes a shooting method for computing three-dimensional equilibria of pre-curved nonlinear beams with axial and shear flexibility using the intrinsic beam formulation. For distributed and concentrated follower loads acting on a cantilevered beam, the method amounts to a direct solution approach requiring only a single shot (zero iterations) to compute the equilibria. This is possible since the system equations are defined in a local coordinate system that rotates and translates with the beam, akin to the follower loads themselves. A general procedure employing nonconservative follower loads, which invokes the Picard–Lindelöf theorem on uniqueness and existence of solutions, is also introduced for finding all solutions for three-dimensional pre-curved beam problems with conservative loading. This is particularly useful in beam buckling problems where multiple stable and unstable solutions exist. Three-dimensional equilibrium solutions are generated for many loading cases and boundary conditions, including three-dimensional helical beams, and are compared to similar solutions where available in the literature. Excellent agreement is documented in all comparison cases. For buckling examples, the stability of the computed solutions is assessed using a dynamic finite element code based on the same intrinsic beam equations. Due to the ability to avoid iteration, the presented approach may find application in model-based control for practical three-dimensional problems such as the control of manipulators utilized in endoscopic surgeries and the control of spacecraft with robotic arms and long cables.     

A differential quadrature procedure for three-dimensional buckling analysis of rectangular plates

This paper presents an application of the differential quadrature (DQ) method for three-dimensional buckling analysis of rectangular plates. The governing equations of the plate model are first presented in terms of displacement, stress displacement relationship, and boundary conditions with three-dimensional flexibility. These equations are then normalised and discretised using the DQ procedure. Example problems pertaining to the buckling of rectangular plates with generic boundary conditions are selected to illustrate the efficiency and simplicity of implementing the DQ procedure. The convergence characteristics of the method are first conducted based on numerical studies. The DQ solutions are then compared, where possible, with exact or approximate solutions. It is found that the differential quadrature method yields accurate results for the plate problems under the current investigation. In addition to the above, some parametric studies are carried out by varying the plates aspect ratio, boundary conditions and thickness to width ratio under axial and biaxial loading.     

离散变量优化设计的方向差商法

本文根据离散变量的特点，提出了方向差商的概念，并且针对一类目标函数、约束函数具有单调性的优化设计问题，发展了一种在可行集外由目标函数最小的点出发，沿目标函数增加最小、约束函数降低最多的方向逐步搜索，逐步靠近可行集边界的方向差商法。并将此方法应用于离散变量结构优化设计中。     

3-D Aerodynamic Shape Design Using Hessians Based on Incomplete Sensitivities

We investigate the practicability of an optimization algorithm based on the Broyden-Fletcher-Goldfarb-Shanno (BFGS) method for 3D shape design problems, using approximate sensitivities of objective functions, where the contribution of the partial derivatives of the flow state with respect to the control variables is neglected. Therefore, it is worthwhile to investigate how an optimization method based on the Hessian behaves in this context. Indeed, the Hessian should be far from its real value if the gradient approximation is wrong. The optimization methodology is characterized by an unstructured CAD-free framework for shape and mesh deformations, an automatic differentiation of programs for the computation of the gradient of the cost function, and an unstructured flow solver. The redesign of transonic and supersonic wings has been considered and the performance of the BFGS method has been analyzed in comparison with a steepest descent method. Taking into account that a line search is too expensive to be carried out in such problems, a step size proportional to the gradient modulus has been employed for updating the control variables. Numerical results show that the BFGS method does not suffer from the approximation used in the evaluation of sensitivities, and leads to an effective improvement of the efficiency of the optimization methodology. These results can be then considered an a posteriori justification for incomplete sensitivities.     

用具有负定非对称矩阵的梯度系统构造稳定的广义Birkhoff系统1）

Birkhoff系统是一类比Hamilton系统更广泛的约束力学系统,可在原子与分子物理,强子物理中找到应用.非定常约束力学系统的稳定性研究是重要而又困难的课题,用构造Lyapunov函数的直接方法来研究稳定性问题有很大难度,其中如何构造Lyapunov函数是永远的开放问题.本文给出一种间接方法,即梯度系统方法.提出一类梯度系统,其矩阵是负定非对称的,这类梯度系统的解可以是稳定的或渐近稳定的.梯度系统特别适合用Lyapunov函数来研究,其中的函数V通常取为Lyapunov函数.列出广义Birkhoff系统的运动方程,广义Birkhoff系统是一类广泛约束力学系统.当其中的附加项取为零时,它成为Birkhoff系统,完整约束系统和非完整约束系统都可纳入该系统.给出广义Birkhoff系统的解可以是稳定的或渐近稳定的条件,进一步利用矩阵为负定非对称的梯度系统构造出一些解为稳定或渐近稳定的广义Birkhoff系统.该方法也适合其他约束力学系统.最后用算例说明结果的应用.     

High accuracy postbuckling analysis of channel section struts

This paper presents the theoretical developments of a high accuracy method for the post-buckling analysis of some channel section struts. In this method, the Von-Karman's equilibrium equation is solved exactly to obtain the buckling loads and the corresponding form of out-of-plane buckling deflection modes. The investigation of channel section buckling behaviour is then extended to the post-buckling study with the assumption that the deflected form after the buckling is the combination of first, second and higher (if required) modes of buckling. Thus, the full-analytical post-buckling study is effectively a multi term analysis, which is attempted by utilizing the so-called semi-energy method. In this method the Von-Karman compatibility equation is used together with a consideration of the total strain energy of the strut. Through the solution of the compatibility equation, the in-plane displacement functions which are themselves related to the Airy stress function are developed in terms of the unknown coefficients in the assumed out-of-plane deflection function. The in-plane and out-of-plane deflection functions are substituted in the total strain energy expressions and the theorem of minimum total potential energy is applied to solve for the unknown coefficients. It is noted that the Classical Plate Theory (CPT) is applied throughout the theoretical developments. Through the comparison of the results and the appropriate discussion, the knowledge of the level of capability of the method is significantly promoted.     

Nonlinear buckling optimization of composite structures considering “worst” shape imperfections

Nonlinear buckling optimization is introduced as a method for doing laminate optimization on generalized composite shell structures exhibiting nonlinear behaviour where the objective is to maximize the buckling load. The method is based on geometrically nonlinear analyses and uses gradient information of the nonlinear buckling load in combination with mathematical programming to solve the problem. Thin-walled optimal laminated structures may have risk of a relatively high sensitivity to geometric imperfections. This is investigated by the concepts of “worst” imperfections and an optimization method to determine the “worst” shape imperfections is presented where the objective is to minimize the buckling load subject to imperfection amplitude constraints. The ability of the nonlinear buckling optimization formulation to solve the laminate problem and determine the “worst” shape imperfections is illustrated by several numerical examples of composite laminated structures and the application of both formulations gives useful insight into the interaction between laminate design and geometric imperfections.     

Bending and stability analysis of gradient elastic beams

The problems of bending and stability of Bernoulli–Euler beams are solved analytically on the basis of a simple linear theory of gradient elasticity with surface energy. The governing equations of equilibrium are obtained by both a combination of the basic equations and a variational statement. The additional boundary conditions are obtained by both variational and weighted residual approaches. Two boundary value problems (one for bending and one for stability) are solved and the gradient elasticity effect on the beam bending response and its critical (buckling) load is assessed for both cases. It is found that beam deflections decrease and buckling load increases for increasing values of the gradient coefficient, while the surface energy effect is small and insignificant for bending and buckling, respectively.     

Analytical solutions for buckling of size-dependent Timoshenko beams

The inconsistences of the higher-order shear resultant expressed in terms of displacement(s) and the complete boundary value problems of structures modeled by the nonlocal strain gradient theory have not been well addressed. This paper develops a size-dependent Timoshenko beam model that considers both the nonlocal effect and strain gradient effect. The variationally consistent boundary conditions corresponding to the equations of motion of Timoshenko beams are reformulated with the aid of the weighted residual method. The complete boundary value problems of nonlocal strain gradient Timoshenko beams undergoing buckling are solved in closed forms. All the possible higher-order boundary conditions induced by the strain gradient are selectively suggested based on the fact that the buckling loads increase with the increasing aspect ratios of beams from the conventional mechanics point of view. Then, motivated by the expression for beams with simply-supported(SS) boundary conditions, some semiempirical formulae are obtained by curve fitting procedures.