复变量平均法与其他近似方法的异同EI北大核心CSCD

复变量平均法因其通用性和实用性受到学界的大量关注,但在求解系统响应时会产生一定误差。该研究旨在通过比较不同近似方法间的区别揭示各方法的精度差异和适用条件。应用复变量平均法、多尺度法和谐波平衡法获得单自由度自治和非自治系统的近似解析解,并以Duffing振子为算例进行数值验证。随后针对二自由度非线性能量阱系统,推导出系统稳态响应的半解析解,以振幅和均方根值为评价指标描述系统的响应情况。结果表明:对于单自由度系统,复变量平均法和多尺度法得到的衰减振动瞬态解相同,不同于谐波平衡法;三种方法获得的受迫振动稳态解相同。三者对于弱非线性自治系统和非自治系统响应的近似准确率较高。复变量平均法和谐波平衡法均能良好地描述二自由度耦合系统的稳态周期运动且精度较高。当出现拟周期运动时,以均方根值为指标,复变量平均法的解析效果更好;以振幅为指标,谐波平衡法的近似程度更高。     

二阶强非线性非自治系统周期解的能量迭代法

给出了确定一般二阶强非线性非自治系统周期解的能量迭代方法。该方法给出了二阶强非线性非自治系统主谐波共振和超谐波、次谐波共振周期解存在的必要条件，求得了这些周期解的近似解析表达式，并且得出了解的稳定性判据。近似解析解表达式由计算机辅助推导，计算程序集推导、计算、数值解以及绘图于一体。例子表明，该方法不仅有效而且结果精度较高。     

一个非线性奇异振子的谐波平衡解

应用谐波平衡法计算了一个恢复力与因变量成反比的非线性振子的近似频率和近似周期解。与Mickens的方法不同,直接求解了非线性奇异二阶微分方程。一阶和二阶谐波解所对应的非线性恢复力的傅立叶级数展开式的系数容易由相应的积分得到。由二阶谐波平衡法得到的非线性代数方程组很容易用符号运算软件求出。得到的一近似频率与精确频率的百分比误差是12.8%,而二阶近似频率与精确频率的百分比误差小于1.28%。与数值方法给出的“精确”周期解比较,二阶近似解析解要比一阶近似解析解精确得多。高阶谐波平衡法一般需要求解复杂的非线性代数方程组,但是借助于Matlab和Mathematica等符号运算软件,这一困难可以得到一定程度的克服.     

Chebyshev多项式在锚链分析中的应用

依据水流作用下锚链的数值分析模型，提出了应用Chebyshev多项式拟合建立锚链张力与锚链顶端位置函数关系的近似方法，该锚链函数多项式可方便地用于波浪与锚泊结构物相互作用的实时模拟中。计算中按静水和有流两种情况，应用二维和三维Chebyshev多项式，建立了锚链顶端水平和垂向拉力与锚链顶端水平、垂向位置，及水流速度的函数关系。锚链分析模型采用分段外推－校正方法计算，在无流、均质锚链情况下，计算结果与解析解完全吻合。     

分数阶单自由度间隙振子的受迫振动

研究了含有分数阶微分项的单自由度间隙振子的受迫振动,利用KBM渐近法获得了系统的近似解析解。分析了分段线性系统的主共振,得到了分数阶阶次在0～2时分数阶项的统一表达式;发现分数阶微分项在分段系统中以等效线性阻尼和等效线性刚度的形式影响着系统的动力学特性,而间隙以等效非线性刚度的形式影响着系统的动力学特性。获得了主共振幅频响应的表达式,并得到了系统的稳定性条件;比较了系统主共振幅频响应的近似解析解和数值解,发现两者符合程度较高,验证了近似解析解的正确性;详细分析了分数阶项和间隙对系统主共振幅频响应的影响。研究表明KBM渐近法是分析分数阶分段光滑系统动力学的有效方法。     

含间隙强非线性扭振系统的分岔行为研究

建立了含间隙旋转机械强非线性扭振系统的动力学方程。应用MLP法求解谐波激励下强非线性系统的解析近似解,并运用MLP法与多尺度法结合的方法得到该系统的分岔响应方程。采用奇异性理论研究了系统在非自治情形下的分岔特性,得到不同参数下系统的分岔形态。最后通过具体算例,利用数值模拟的方法得到系统在强非线性项参数变化下的分岔行为,发现随着系统参数变化系统发生周期运动、倍周期运动以及混沌等多种运动形态的复杂动力学行为。研究结果为分析间隙引起的旋转机械传动系统扭振特性提供一定的理论指导和参考。     

简支半球壳的轴对称弯曲

本文用三维弹性力学理论求得简支半球壳轴对称弯曲问题的一般解析解。该解具有分离变量形式，在坐标方向是正交完备的雅可比多项式级数。文末结合一个算例给出数字结果，并证实了中厚壳理论的精确性。     

一类非线性jerk方程的改进两变量展开法

应用改进的两变量展开法求解非线性含有三次非线性项的三阶微分方程的近似频率和近似解析周期解。该方法结合了Lindstedt-Poincare方法与两变量展开法不仅可以适用于弱非线性振动问题的求解而且还可以适用于强非线性振动问题的求解。文中以一个不含速度线性项的非线性jerk方程作为例子分析并得到二阶近似周期和二阶近似解析周期解,与数值方法给出的“精确”周期解比较,二阶近似解析周期解比一阶近似解析周期解要精确得多。结果表明,改进的两变量展开法能够适用于求解非线性jerk方程。而且在jerk方程不含速度线性项时该方法仍然有效。     

一种新的非高斯随机振动疲劳寿命估计方法

构造了非高斯修正系数的多项式响应面模型,提出了一种基于高斯近似的非高斯随机振动疲劳寿命估计方法。采用Winterstein传递函数法将非高斯随机应力转化成高斯随机应力,并联合雨流计数和Miner损伤准则分别估算两种随机应力下的累计损伤和谱矩,对多个样本进行最小二乘拟合之后构建一个关于应力谱矩和非高斯修正系数的多项式响应面模型。利用高斯近似法估算非高斯随机振动疲劳损伤量,并与经过雨流计数和Miner损伤准则估算的非高斯随机过程下疲劳损伤对比,结果表明:高斯近似法具有较好的精度。     

一类强非线性振动系统的改进能量解析法

能量法是求解强非线性系统常用的一种近似解析方法,由于忽略了激励力和阻尼力所做的瞬时功的影响,计算上比较方便。但是当扰力水平较高时,能量解法往往会产生较大误差。揭示了能量法产生误差的原因,并对传统的能量法进行了修正,提出了一种新的解析法-改进能量法。以动力基础为研究对象,用改进能量法给出了此类强非线性非自治系统的解析解,并用Runge-Kutta法对该系统进行了数值计算。分析表明,改进能量法比传统的能量法有了很大的改进,即便是在扰力水平较高的亚谐振动条件下,仍然能够给出较为准确的解。     

单摆大振幅振动的解析逼近解

构造了单摆大振幅振动的高精度解析逼近周期和周期解.首先,利用Maclaurin展开和Chebyshev多项式加速技术将单摆振动方程中的正弦型恢复力用三次多项式近似代替,得到一个Duffing型方程;然后,将牛顿法与谐波平衡法结合起来建立Duffing方程的解析逼近周期及周期解,从而给出单摆振动的解析逼近解.因此,在求解过程中避免了关于参数的非线性代数方程组的出现,只需解线性代数方程组就能建立单摆振动的解析逼近周期及周期解.几乎在振幅(初始摆角)的全部取值范围内,与数值方法给出"精确"周期及周期解比较,得到的解析逼近解都有很高的逼近精度.     

Analysis of nonlinear heat conduction based on determining the front of temperature perturbation

An analytical solution of nonlinear problem of heat conduction is derived using an integral method of heat balance. In order to improve the accuracy of solution, the temperature function is approximated by polynomials of higher degrees. The polynomial coefficients are determined using additional boundary conditions which are found from the basic differential equation and preassigned boundary conditions including the conditions on the front of temperature perturbation. It is demonstrated that the introduction of additional boundary conditions even in a second approximation results in a significant increase in the accuracy of solution of the problem.     

解析型Winkler弹性地基梁单元构造

该文采用Winkler弹性地基梁理论确定了弹性地基梁的挠度方程解析通解; 根据最小势能原理建立了解析型Winkler弹性地基欧拉梁及铁摩辛柯梁的单元刚度及等效节点荷载; 得到了解析型弹性地基欧拉梁单元AWFB-E及铁摩辛柯梁单元AWFB-T。同时,论文还采用传统里兹法求得了相应的Winkler弹性地基欧拉梁及铁摩辛柯梁单元刚度矩阵,得到了里兹法弹性地基欧拉梁单元RWFB-E及铁摩辛柯梁单元RWFB-T。对该文构建的两类单元与一般梁-基体系有限元分析结果及理论解析解进行了对比。对比结果表明,传统里兹法由于其多项式形函数无法精确模拟弹性地基梁变形,因此其结果与理论解析解有误差,但随着单元数量增多其误差减小; 采用解析型单元进行计算时,无论单元数量多少,得到的均为“真实”解,说明解析试函数法求得的位移形函数比一般的多项式形函数精确,得到的弹性地基梁单元具备解析型、精确性的特点,可应用于解决实际工程问题。     

Accurate approximation to the double sine-Gordon equation

In general, this paper deals with nonlinear double sine-Gordon equation with even potential energy which has arisen in many physical phenomena. The nonlinear dispersion problems without a small perturbation parameter are difficult to be solved analytically. Hence, the main concern is focused on solving the traveling wave of the double sine-Gordon equation. As commonly known, the perturbation method is for solving problems with small parameters, and the analytical representation thus derived has, in most cases, a small range of validity. For some nonlinear problems, although an exact analytical solution can be achieved, they often appear in terms of sophisticated implicit functions, and are not convenient for application. Although a variety of transformation methods has been developed for solving the nonlinear dispersion problems, such transformed equations still include nonlinear terms. To overcome these difficulties, a new approach for Newton-harmonic balance (NHB) method with Fourier–Bessel series is presented here. It is applied to solve the higher-order analytical approximations for dispersion relation in double sine-Gordon equation. The Fourier–Bessel series with the NHB method presents excellent improvement from lower-order to higher-order analytical approximations involving the nonlinear terms in the double sine-Gordon equation. Not restricted by the existence of a small perturbation parameter, the method is suitable for small as well as large amplitudes of wavetrains. Excellent agreement with exact solutions is presented in some practical examples.     

谐变力作用功能梯度旋转圆板强非线性主共振

功能梯度材料作为一种新型材料,具有良好的力学性能,近年来被广泛关注和应用。该文针对金属-陶瓷功能梯度圆板,考虑周边夹支边界约束条件,选取多项式形式的振型函数,利用伽辽金法,推得旋转运动状态和热效应作用下系统的纵横耦合非线性振动方程,求得由旋转及密度差引起的静挠度项。用改进多尺度法求解方程,得到强非线性系统的频幅响应方程和解析解。通过算例,给出功能梯度圆板的幅频曲线、幅值-激励力曲线、幅值-温度曲线,分析了不同物理量对结构共振幅值的影响规律,并且比较了解析解和数值解,两者结果较为吻合。     

Analytical approximations to the solutions for a generalized oscillator with strong nonlinear terms

This paper is focused on solving the generalized second-order strongly nonlinear differential equation ${\ddot{x}+\sum_i {c_i^2 }x \left| x \right|^{i-1}=0}$ which describes the motion of a conservative oscillator with restoring force of series type with integer and noninteger displacement functions. The approximate analytical solution procedures are modified versions of the simple solution approach, the energy balance method, and the frequency?Camplitude formulation including the Petrov?CGalerkin approach. For the case where the linear term is dominant in comparison with the other series terms of the restoring force, the perturbation method based on the solution of the linear differential equation is applied. If the dominant term is nonlinear and the additional terms in the restoring force are small, the perturbation method based on the approximate solution of the pure nonlinear differential equation is introduced. Using the aforementioned methods, the frequency?Camplitude relations in the first approximation are obtained. The suggested solution methods are compared and their advantages and disadvantages discussed. A numerical example is considered, where the restoring force of the oscillator contains a linear and also a noninteger order term (i?=?5/3). The analytically obtained results are compared with numerical results as well as with some approximate analytical results for special cases from the literature.     

A multigrid method for the generalized symmetric eigenvalue problem: Part I—algorithm and implementation

A multigrid method is described that can solve the generalized eigenvalue problem encountered in structural dynamics. The algorithm combines relaxation on a fine mesh with the solution of a singular equation on a coarse mesh. A sequence of coarser meshes may be used to quickly solve this singular equation using another multigrid method. The hierarchy of increasingly finer meshes can be further exploited using a nested iteration scheme, whereby initial approximations to the fine mesh eigenvectors are computed using interpolated coarse mesh eigenvectors. The solution of some simple plate problems on a Convex C240 demonstrates the efficiency of a vectorized version of the multigrid algorithm.     

A multigrid scheme for the implicit solution of the compressible flow equations

A multigrid scheme has been developed for the acceleration of the solution of compressible inviscid and viscous flow problems. A higher order accurate upwind conservative finite volume scheme has been used for the discretization of the Euler and the Reynolds-Averaged Navier-Stokes equations. For the multigrid implementation, the alternative point of view of the Full Approximation scheme has been employed together with a conservative restriction operator to maintain the fine grid accuracy. The present multigrid scheme has been designed to take full advantage of the implicit unfactored solution scheme of the single grid code by introducing an alternative multigrid V-cycle. The proposed method attains up to 25-fold acceleration with respect to the single grid solution for moderate size grids. Moreover, the results demonstrate that the computational time increases proportional to the number of volumes when global refinements are applied so, the present multigrid scheme is very favorable for large scale computations.This work was partially supported by the EC-programme: BRITE/ EURAM Project Aero 0018. The authors would like to thank MBB GmbH for providing NsFlex code