压电传感器探测层合板分层的数值分析

用修正的准三维有限元法分析了受单轴拉伸表面覆盖有压电薄膜的复合材料层合板。文中首先讨论了压电薄膜层对层合板层间应力的影响，然后在存在分层裂纹的情况下，计算了压电层的电压分布。结果表明将压电薄膜作为传感器来进行分层探测是可行的。文中最后考虑了多种因素对传感灵敏度的影响。     

斜拉索各参数取值对索力测定结果的影响

为了确保频率法测定斜拉索索力的精度，以某斜拉桥中的拉索为例，采用平面非线性有限元法探讨了斜拉索各参数取值对索力值的影响，并讨论了各参数的合理取值．利用分析结果找到了不同的测试单位在该斜拉桥的索力联测中存在较大索力偏差的原因，可对施工过程中的索力精确测定提供帮助．分析结果同样适用于悬索桥和悬挂屋顶等索结构的索力测定．     

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。     

Numerical Investigation of Large Elastoplastic Strains of Three-Dimensional Bodies

A method of stress—strain analysis of elastoplastic bodies with large displacements, rotations, and finite strains is developed. The incremental loading technique is used within the framework of the arbitrary Lagrangian—Eulerian formulation. Constitutive equations are derived which relate the Jaumann derivative of the Cauchy—Euler stress tensor and the strain rate. The spatial discretization is based on the FEM and multilinear three-dimensional isoparametric approximation. An algorithm of stress—strain analysis of elastic, hyperelastic, and perfectly plastic bodies is given. Numerical examples demonstrate the capabilities of the method and its software implementation __________ Translated from Prikladnaya Mekhanika, Vol. 41, No. 6, pp. 36–43, June 2005.     

透平机械离心压缩机和离心泵叶轮动力分析的三维旋转循环对称CN群算法

以离心压缩机叶轮为例,建立三维有限元模型,采用循环对称CN群算法计算了它的动力特性,详细讨论离心力作用下叶轮的应力状态,分析与轴连接的不同方式对叶轮应力的影响。应用循环对称算法能大大减少建模与计算的工作量,有很好的实用价值。计算结果表明,离心压缩机叶轮在旋转中,叶轮外缘应力比较高,轮盖上的应力较轮盘上的应力大,叶片根部容易产生应力集中,应采用适当的园角避免应力集中,过盈配合的叶轮内孔处应力比较大,需要确定合理的过盈量,既要保证叶轮不松动也要保证叶轮内孔处的强度条件。     

The effective properties of composites with fiber-end cracks

The paper describes use of self-consistent finite element method (SCFEM) for predicting effective properties of fiber composite with partially debonded interface. The effective longitudinal Young's modulus and shear modulus for unidirectional fiber reinforced composites with fiber-end cracks are calculated. Numerical results show that the effective properties are considerably influenced by the fiber-end cracks. The effects of microstructural parameters, such as fiber volume fraction, modulus ratio of the constituents and fiber aspect, on the effective properties of the composites were discussed. The project supported by the National Natural Science Foundation of China     

光纤传感器对机敏复合材料结构性能的影响

本文通过实验研究了光纤传感器对埋光纤机敏复合材料机械性能的影响，详细分析了埋光纤机敏复合材料层板在外载作用下的应力分布，针对三种不同铺层方式，研究了六种树脂填充区对机敏复合材料层板应力集中的影响，得出了减小应力集中的有效方案。     

Analysis of blank thickening in deep drawing processes using the theory of a Cosserat generalized membrane

This paper analyzes a transient, nonlinear deep drawing process where a circular blank of a rigid-plastic material is forced by a rigid circular punch to deform into a cylindrical cup. Attention is focused on the plastic flow beneath the blank-holder. Using the Cosserat theory of a generalized membrane it is possible to obtain analytical solutions which examine the following two major effects: (a) the importance of added “rim pressure” acting on the outer edge of the blank; and (b) the importance of a controlled moveable blank-holder to allow blank thickening during the drawing process. Guided by these analytical results, a new deep drawing machine was built to exploit these effects and increase the limit drawing ratio (LDR) of the drawing process. Specifically, the LDR (in one stroke) reached the value of 3.16 compared with the value of about 2.0 in the conventional process. Moreover, the analytical prediction of the punch force versus the punch stroke is in good agreement with the experimental data and with simulations using the computer code DYTRAN.     

矩形板结构损伤的分区域神经网络识别方法

通过引入LM优化算法，针对矩形薄板中对称结构的损伤识别问题，提 出了一种基于神经网络的分区域分步识别方法. 对于预测输出量比较多且对预测精度要求比 较高的问题，常会出现网络训练时收敛速度慢、网络预测精度低，并且当网络训练达到目标 误差时，输出的预测量中常有某个输出量的误差还很大的情况. 针对这些问题，利用选 取的组合输入参数，提出了基于神经网络的分区域识别方法. 通过对悬臂板结构的数值模拟 结果表明：提出的分区域识别方法对结构损伤的分区和预测是可行和有效的， 其预测精度要明显的高于只用单个网络的预测结果，并且预测子网络对损伤的位置和程度是 同步输出的，从而避免了传统分步识别理论中子网络过多的问题.     

某边坡变形破坏机理的数值分析

某花岗岩矿岩土体主要有填土、粉质粘土、淤泥及淤泥质土、含泥中粗砂、粘土、残积砂质粘土,其下为花岗岩基岩。边坡最大高度75m,其中残积土层以上的土质边坡高度约50m。本文运用有限元强度折减法对该边坡的稳定性进行了研究,得出了潜在的破坏模式及各个滑动部分的安全系数。分析结果表明,本边坡存在3个潜在滑动面,其中两个为浅层滑动面,另一个为深层滑动面。两个浅层滑动面的稳定系数在1.13～1.17,处于临界状态。深层滑动面稳定性系数在1.29,目前尚处在相对稳定状态。根据分析结果,提出了如下加固措施:（1）对滑坡体进行削坡处理;（2）在其滑动面下方设置抗滑台阶;（3）采用预应力锚索框架梁结合肋柱进行加固处理。