高速滚动轴承-转子系统时变轴承刚度及振动响应分析

高速滚动轴承广泛应用于机床主轴、航空发动机等转子系统中。在复杂运行工况下,滚动轴承的刚度表现出强烈的时变特性和非线性特性,往往是系统非线性的主要根源。考虑离心力、陀螺力矩、轴承内圈离心膨胀和热变形等因素,建立高速滚动轴承力学模型,计算轴承的时变刚度。将滚动轴承非线性模型与转子有限元模型集成,建立滚动轴承-转子耦合系统动力学模型。以FAG角接触球轴承(HCB7012E)为例,分别计算静载荷作用下的内外圈轴向、径向相对位移,并与舍弗勒轴承分析软件BearinX?的计算结果进行比较,验证了模型对静态位移仿真的精度。在不同轴承预紧状态下,仿真滚动轴承-转子系统在不平衡激励下的振动响应,并与试验结果比较,验证了模型仿真系统动态响应的精度。利用一个背对背安装的角接触球轴承-转子系统,研究在静载荷、不平衡载荷激励作用下滚动轴承刚度的变化规律,并计算时变轴承刚度作用下转子的时域振动响应及频域特征,为高速滚动轴承-转子系统设计、动力学分析与故障诊断提供依据。     

基于非线性理论的齿轮机构动力学模型的建立及实验

在研究分析了轮齿的变啮合刚度和轮齿间的油膜动力学性能的基础上，提出了把此油膜看成是一个质量-弹簧-阻尼系统的假设，并运用弹流理论，分析了油膜厚度和油膜等效刚度，建立了相应的动力学模型和运动微分方程。采用GEAR方法求解非线性动力学微分方程组，并通过实验验证齿侧间隙对齿轮机构动态特性的影响。     

基于改进深度卷积神经网络的步态识别算法

针对传统步态识别算法因服饰携带物变换、视角等协变量因素变化导致的识别能力下降问题,提出了一种基于改进深度卷积神经网络的步态识别算法。该算法利用分层处理机制从步态数据中提取步态特征,能够降低常见变化和遮挡等情况对识别精度的影响,同时,算法根据实验确定了网络中每层特征图的最佳数量、特征图的最佳尺寸以及要用于步态识别的输入特征的类型,能够处理相对较小的数据集而无需使用任何增强或微调技术。CASIA-B步态数据库仿真实验表明,所提出的卷积神经网络覆盖了交叉视图步态识别和无主题的步态识别问题,能够克服与步态识别相关的协变量因素问题,具有更优的步态识别精度。     

A direct complementarity approach for the elastoplastic analysis of plane stress and plane strain structures

This paper presents a direct complementarity approach for carrying out the elastoplastic analysis of plane stress and plane strain structures. Founded on a traditional finite‐step formulation, our approach, however, avoids the typically cumbersome implementation of iterative predictor–corrector procedures associated with the ubiquitous return mapping algorithm. Instead, at each predefined step, the governing formulation—cast in its most natural mathematical programming format known as a mixed complementarity problem—is directly solved by using a complementarity solver run from within a mathematical modeling system. We have chosen the industry‐standard General Algebraic Modeling System/PATH mixed complementarity problem solver that is called from within the General Algebraic Modeling System environment. We consider both von Mises and Tresca materials, with perfect or hardening (kinematic and isotropic) behaviors. Our numerical tests, five (benchmark) examples of which are presented in this paper, have been carried out using models constructed from the mixed finite element of Capsoni and Corradi (Comput. Methods Appl. Mech. Eng. 1997; 141 :67–93), which beneficially offers a locking‐free behavior and coarse‐mesh accuracy. The results indicate, in addition to an isochoric locking‐free behavior, good accuracy and the ability to circumvent the difficult singularity problem associated with the corners of Tresca yield surfaces. Copyright © 2012 John Wiley & Sons, Ltd.     

HP40Nb合金钢高温劣化的非线性超声评价

耐高温合金材料在服役过程中会发生高温劣化,主要表现为微观组织结构的变化,从而导致在其中传播的超声波能产生显著的高次谐波.针对HP40Nb合金钢材料高温劣化的检测及评价,提出采用归一化超声纵波非线性参量来表征材料的高温损伤状态.对HP40Nb损伤试样进行了非线性超声信号的测量结果显示,归一化超声纵波的非线性参量随着材料高温加载时间呈现出显著的上升-平稳-下降趋势.对材料微观组织演化过程进行了扫描电镜(SEM)的观察和分析,结果表明,超声非线性参量在早期阶段的上升与HP40Nb材料高温加载过程中η相等第二相的析出和聚集密切相关;其在后期阶段的下降归因于析出相的减少和脱落以及微孔洞的产生.由此可见,非线性超声对HP40Nb材料高温劣化后微观组织的变化非常敏感.     

Efficient and flexible protocol for implementing two-qubit controlled phase gates with cross-Kerr nonlinearity

The controlled phase gate is one of the most important logic gates in the quantum computation field. In this paper, we proposed a protocol for implementing the two-qubit controlled phase gates with the help of cross-Kerr nonlinearity, optical elements and the conventional photon detectors, which are feasible with existing experimental technology. The protocol also can be applied to implement the controlled phase gates of many different atomic and photonic degrees of freedom with successful probability of 100%, that is, our protocol is efficient and flexible.     

Modal-based damage identification for the nonlinear model of modern wind turbine blade

In this paper, the modal-based indices are used in damage identification of the wind turbine blade. In contrast of many of previous researches, the geometric nonlinearity due to the large structural deformation of the modern wind turbines blade is considered. In the first step, the finite element model (FEM) of the rotating blade is solved to obtain the modal features of the deformed structure under operational aerodynamic loading. Next, the accuracy and efficiency of the various modal-based damage indices including the frequency, mode shape, curvature of mode shape, modal assurance, modal strain energy (MSE) and the difference of indices (between the intact and damaged blades) are investigated. To adapt the MSE index calculation in nonlinear modeling, a new approach is introduced to include the effects of the structural nonlinearity. Furthermore, the effect of the damage length, its location and severity and also the effect of rotational speed and amplitude of loading are studied. The generic 5-MW NREL blade is used for the simulation study. The results show enough sensitivity of the mode shape curvature and MSE indices to the local damages. Moreover, the importance of geometric nonlinearity in the damage detection of the modern wind turbines is demonstrated.     

一种压电半导体纳米线的热电耦合性能研究

采用有限元分析方法,研究了一种n型压电半导体纳米线(氧化锌)的电热耦合性能,分析了外部温度对氧化锌纳米线内部机械场、电场及电流场分布的影响,并讨论了本构方程线性化对电学参数的影响。研究结果表明,温度对氧化锌纳米线的电场、载流子浓度和电流密度影响很大,采用线性本构和非线性本构求得的电场、电子浓度和电流密度最大相差分别为24%,32%和68%,基于非线性本构分析压电半导体的电学性能会引起很大误差。该研究结果可为压电半导体器件利用温度调控电场、电流提供理论依据。