汽车制造中的焊接变形预测

通过完成的几个典型汽车结构件实例，介绍了焊接变形预测在汽车制造中的应用和特点。研究表明，由于汽车部件属于薄壁结构，采用板壳单元固有应变法预测焊接变形更加实用。预测结果与实测数据吻合较好，证明焊接变形预测可以成功地应用于汽车制造。研究结果为实际生产提供了参考依据。     

焊接变形预测与控制的研究进展

综述了焊接变形预测和控制的研究现状，介绍了对焊接变形进行预测的几种常用理论和方法，包括热弹塑性理论、考虑相变和耦合效应的有限元分析、粘弹塑性有限元分析、残余塑变法、相似理论、人工神经网络的应用等。同时介绍了控制焊接变形的几种方法，如反变形法、控制工艺参数、控制温度场以及系统综合分析等。最后讨论了焊接变形研究中需要解决的几个问题。     

基于BP神经网络的齿圈装夹变形预测研究

针对薄壁齿圈的装夹变形问题,将Abaqus有限元仿真与BP神经网络技术应用到了齿圈装夹变形预测中。根据齿圈实际加工装夹情况,应用Abaqus有限元分析软件,建立了齿圈装夹变形的仿真模型,开展了齿圈装夹变形的有限元分析研究,建立了齿圈装夹力及其径向最大装夹变形之间的关系;以Abaqus有限元仿真数据作为训练样本和检验样本,借助BP神经网络良好的预测精度和非线性泛化能力,通过MATLAB神经网络工具箱,建立了基于BP神经网络的齿圈装夹变形预测数字化模型;并根据检验样本对模型进行了检验,预测值与仿真值之间的相对误差在0.05%之内。研究结果表明:建立的基于BP神经网络的齿圈装夹变形预测数字化模型是准确有效的,可以为智能化大数据加工制造环境下的齿圈装夹参数优化提供准确有效的数据。     

基于机器视觉的正交微切屑变形系数研究

切屑变形系数是描述切削变形程度的重要参数,是计算其它切削过程参数的基础。然而宏观的切削理论和试验技术不适用于微切削技术。研究微切屑变形系数对预测加工结果、优化切削过程、控制加工参数有着重要意义。本文应用机器视觉方法获得了微切削中切屑的几何形状,找出了切削厚度、厚度变形系数、宽度变形系数与进给量、背吃刀量的变化关系,为微切削过程的控制和预测提供了理论依据。试验方法简单、精确、效率高。     

焊接变形预测的研究进展

焊接变形预测技术对焊接结构的制造和使用有重要意义,数值模拟是近年来研究焊接变形的主要手段和方法,文中综述了预测焊接变形的各种方法及其发展现状,重点介绍了热弹塑性有限元法和固有应变法预测焊接变形的原理、存在的困难和国内外研究者提出的相应解决措施,以及下一步研究有待解决的问题。     

平板对接接头焊接变形的数值模拟与试验研究

为验证数值模拟方法在构件焊接变形预测应用中的可靠性,对某多层焊平板对接接头进行了数值模拟研究与试验验证。共进行了三组焊接试验,采用三维扫描仪对试验后的焊件进行测量,并分析其焊后的残余变形量。基于SYSWELD分析软件,应用热弹塑性有限元法对该平板接头进行了数值模拟。通过将数值模拟仿真结果与试验测量结果进行对比分析,证明了数值模拟方法在焊接变形预测中的可靠性。其中角变形误差为7.8%,横向收缩变形误差为23.3%,纵向收缩变形误差为15.2%。     

基于初始变形的薄板堆焊变形预测技术

针对薄板初始变形对焊接结构件尺寸精度的影响，提出了基于数字图像相关技术的光学检测方法，对薄板堆焊变形预测精度进行全场动态研究。首先采用结构光扫描的方法，获取焊接结构件初始变形轮廓，然后基于初始变形模型进行焊接变形预测，最后通过理想模型和初始变形模型预测数据对比，研究结构初始变形对堆焊变形精度的影响。研究结果表明：基于数字图像相关技术的测量方法是静态、动态获取焊接变形数据的有效手段；相对于理想结构模型，基于初始变形结构的模型，在单点动态和全场静态变形中具有较高的预测精度；两种模型在焊接完成后均呈马鞍形，但初始变形促进了焊接演变过程；初始变形缺陷影响了纵向残余压应力分布，是引起不同焊接变形的根本原因。     

固有应变值施加方式的焊接变形仿真研究

为提高固有应变法在预测焊接变形应用中的准确度,以T型接头为算例,采用SYSWELD仿真软件对多个方案进行数值模拟,研究了固有应变施加方式对焊接变形预测的影响,进而提出了一种新的固有应变施加方式。为了验证此方法的准确性,将三种设计方案的计算结果与热弹塑性有限元方法的计算结果进行分析比较。结果表明:在以焊缝中间线为轴心、焊缝直角边的1/2为半径的范围作为固有应变的施加区域时,其变形预测精度明显提高,变形量更加逼近于热弹塑性有限元分析方法的预测结果。     

基于B样条的航空薄壁件加工变形预测与控制

针对开发航空薄壁件多点柔性工装系统的需求,研究了航空薄壁件多点柔性数控加工中的变形预测与控制问题。研究了利用有限元分析和Matlab编程建立B样条加工变形数学模型并进行预测的方法及高速铣削切削参数在线优化控制加工变形的原理和方法。研究结果表明:基于B样条曲面数学模型可对加工变形进行预测,切削参数的在线优化能够有效地控制加工变形。研究结果为有效地解决柔性工装的设计和制造提供了必要依据。     

盾构施工引起地表变形预测系统的设计与实现

讨论了盾构施工引起地表变形的因素和机理,介绍了采用BP神经网络技术对地表变形预测系统进行设计与实现的过程。结合工程实践,对系统预测精度进行研究,结果表明,采用BP神经网络对盾构施工引起地表变形进行预测是一种行之有效的方法。     

Calculated Experimental Evaluation of the Operating Life of Basic Locomotive Parts for Ensuring Their Safe Operation

The calculated experimental procedures for determining the operating life of load-bearing locomotive structures are presented in a probabilistic form for evaluating and predicting their technical condition to provide grounds for the safe operation of locomotives. The procedures encompass lowcycle (in an elastoplastic deformation) and multicycle (in elastic deformations in the range of at least 10⁷ cycles) loading of parts.     

Analysis of the acoustic emission technique used in laboratory tests of reinforced-concrete structures

Deformation, crack generation, and fracture of reinforced-concrete structures are studied experimentally. Acoustic emission is used as one of the testing techniques to monitor the development of these processes. The performed experimental studies confirm the possibility of using the acousticemission technique to study the accumulation kinetics of flaws in reinforced-concrete structures and theoretically predicting fracturing when it is still difficult to do so on the basis of macroscopic observations. It is established that laboratory acoustic-emission tests of reinforced-concrete samples provide the necessary experimental basis for developing the theory of the deformation and fracture of reinforced concrete.     

Material testing by plotting total deformation curves

The paper deals with the theoretical and experimental aspects of the study of the behavior of materials at the stage of supercritical deformation or work softening, consideration of which is important for the assessment of the durability of structures. The study results show the theoretical confirmation of the effect of the specimen configuration on the possibility of plotting the descending portions of deformation curves. It is shown that the equilibrium supercritical deformation of a material is possible if the stiffness of the loading system is sufficient. This is confirmed by the results of the uniaxial tension tests with unloading at various stages of elastoplastic and supercritical deformation. The method for interpreting experimental data obtained for the specimens of various dimensions under conditions of deformation localization are considered.     

成形刚性筋变形力系数的试验确定

刚性筋在各种机械零件中应用很广,其成形力的确定方法陈旧.为此,用试验的方法,对当前变形力计算公式中的系数Kp进行多种情况试验,得出相应的数据,给出应用图表.并结合实例进行验证,其差值为10%左右.     

基于多翼型特征的非奇异变形感知方法研究

针对传统变形感知方法在复杂翼型结构中常见的病态、奇异等问题，提出了一种基于多翼型特征的非奇异变形重构模型。依据Timoshenko梁变形理论，采用依存插值技术离散单元位移场，建立理论截面应变与测量应变的最小二乘变分函数，推导单元节点变形与测量应变的积分重构模型。该模型的位置无关性有效消除评估截面选取不当引起的奇异，增强重构模型在复杂翼型结构中的适用性。同时，针对应变传感器服役期间常见的环境扰动，以重构精度与鲁棒性为评估指标，建立自适应多目标粒子群优化模型。实验结果表明，提出的重构模型整体测量精度较高，在机翼变形量小于20 mm范围内最大绝对误差为0.26 mm,最大相对均方根误差为0.42%;当变形量增大时，绝对误差随之增大，但相对均方根误差不超过3.5%。因此基于多翼型特征的非奇异变形重构模型能够满足机翼实时重构需求，有效扩展变形感知方法在复杂结构中的应用价值。     

An improved algorithm for McDowell’s analytical model of residual stress

The analytical model for two-dimensional elastoplastic rolling/sliding contact proposed by McDowell is an important tool for predicting residual stress in rolling/sliding processes. In application of the model, a problem of low predicting precision near the surface layer of the component is found. According to the volumeconstancy of plastic deformation, an improved algorithm for McDowell’s model is proposed in order to improve its predicting accuracy of the surface residual stress. In the algorithm, a relationship between three normal stresses perpendicular to each other at any point within the component is derived, and the relationship is applied to McDowell’s model. Meanwhile, an unnecessary hypothesis proposed by McDowell can be eliminated to make the model more reasonable. The simulation results show that the surface residual stress predicted by modified method is much closer to the FEM results than the results predicted by McDowell’s model under the same simulation conditions.     

Spline finite strip analysis of the buckling and vibration of composite prismatic plate structures

A spline finite strip capability is described for predicting the buckling stresses and natural frequencies of vibration of prismatic plate structures which may be of composite laminated construction with arbitrary lay-ups. The plate structures may have general boundary conditions. The capability embraces analyses based on the use of first-order shear deformation plate theory and of classical plate theory, and utilizes substructuring procedures which include the use of superstrips. The theoretical development is not detailed since the present paper reports a very direct extension of a theoretical study developed for the analysis of single plates in an earlier paper in this Journal. A considerable range of buckling and vibration applications is documented and comparison of spline finite strip numerical values of buckling stresses and frequencies is made with results generated using the semi-analytical finite strip method and, in some cases, the finite element method. Buckled and vibrational mode shapes are presented for some applications.     

Plastic deformation analysis in machining of Inconel-718 nickel-base superalloy using both experimental and numerical methods

Surface region plastic deformation of Inconel-718 nickel-base superalloy workpieces was evaluated when machined under orthogonal cutting conditions at various cutting speeds. Plastic deformation analysis was accomplished by determining the residual stress and plastic strain distributions in the surface region. The residual stresses were tensile and maximum near the surface and decreased in magnitude with an increase in depth beneath the machined surface. Similarly, the plastic strains were maximum near the surface and decreased with an increase in depth beneath the machined surface. In addition, a finite element simulation of orthogonal machining was carried out for predicting the residual stress and plastic strain distribution. In general, the trend of the curves predicted by the finite element model was similar to those found experimentally.     

计算机辅助质谱预测和候选结构评价

本文介绍一个用基本的质谱裂解规则来预测已知结构化合物的质谱的程序。如果给定对应某未知物的一组候选分子结构,该程序可通过预测质谱的比较来评价这些候选结构。程序采用动态描述方法存储裂解规则,可自动进行各种结构类型化合物的质谱预测。     

基于BP神经网络的TC4钛合金超塑性变形后组织及性能预测研究

分别用Visual Fortran语言和MATLAB软件建立了TC4钛合金超塑性变形时变形参数与其力学性能和晶粒尺寸之间的BP神经N络模型，通过用较少的力学性能和晶粒尺寸的试验数据进行训练，进而对其性能进行预测。结果表明，BP神经网络用于材料超塑性变形后的力学性能及晶粒尺寸预测是可行的，其预测误差小于7%。