旋转电弧角焊工艺参数优化及热应力数值模拟

发布时间：2018-08-11 13:43

【摘要】：旋转电弧角焊是焊接领域中的重要工艺,不同的焊接参数对应的焊接温度场与应力应变场又有所不同,对于焊接后续变形控制研究造成较大阻碍,因此有必要利用焊接智能优化算法来优化焊接工艺参数,并与数值模拟方法相结合,从而得到高质量成形焊缝,而较优工艺参数又能够用于焊接热力学研究,这样节省试验成本,缩短研发周期。为对旋转电弧角焊进行焊接工艺参数优化,减少试验次数,提高焊接效率,首先采用基于minitab的正交试验法,并充分考虑各焊接参数之间相互作用与影响,建立正交试验模型并进行相应的焊接试验,然后测量焊后各工件角焊缝尺寸,利用主观分析法与残差分析法分析该正交模型中各焊接工艺参数间相互匹配关系与拟合优度问题。为快速获得较优焊接工艺参数,基于BP(Back Propagation,BP)神经网络建立旋转电弧角焊焊缝尺寸预测模型。通过对预测样本数据的训练,得出焊接参数与焊缝尺寸之间的映射关系,其预测样本可验证调试后的遗传神经网络。基于BP神经网络非线性映射预测能力与遗传算法全局寻优能力,建立遗传神经网络的旋转电弧焊接参数优化模型,对焊接参数进行了优化。为准确模拟工件在焊接过程中所受外载荷,采用基于ANSYS的数值模拟与试验相结合的方法,以位移约束和集中力约束的施加与释放模拟焊接夹具的外拘束,通过接触分析得到工件所承受的支持力与夹紧力。最后,依据所设计的试验平台分别建立3种不同的多体耦合模型模拟夹具与工件的夹紧作用,分析工件上、下侧板von Mises应力与X、Y向应力分布规律及X、Y向变形形成机理。由于外拘束模型仿真分析结果需试验验证,为此实验室搭建了旋转电弧角焊工件变形测量试验平台,采用角焊专用量规对工件下侧板对应测量点进行了准确测量以得到工件角变形与弯曲变形,进而对旋转电弧角焊热应力数值模拟平台进行完善与改进。
[Abstract]:Rotating arc angle welding is an important technology in welding field. The welding temperature field and stress-strain field corresponding to different welding parameters are different, which is a big obstacle to the research of welding subsequent deformation control. Therefore, it is necessary to optimize welding parameters by using welding intelligent optimization algorithm, and combine with numerical simulation method to obtain high quality formed weld, and the better process parameters can be used in the study of welding thermodynamics. This saves the test cost, shortens the research and development cycle. In order to optimize the welding process parameters, reduce the number of tests and improve the welding efficiency, the orthogonal test method based on minitab is adopted, and the interaction and influence of the welding parameters are fully considered. The orthogonal test model was established and the corresponding welding test was carried out. Then the angle weld size of each workpiece after welding was measured. The matching relationship and the goodness of fit among the welding process parameters in the orthogonal model were analyzed by the subjective analysis method and the residual analysis method. In order to obtain better welding parameters quickly, a prediction model of welding seam size for rotating arc angle welding was established based on BP (Back Propagation BP (BP) neural network. The mapping relationship between welding parameters and weld size is obtained by training the predicted sample data. The predicted sample can verify the genetic neural network after debugging. Based on BP neural network nonlinear mapping prediction ability and genetic algorithm global optimization ability, the optimization model of rotating arc welding parameters based on genetic neural network is established, and the welding parameters are optimized. In order to accurately simulate the external load of workpiece in welding process, the method of combining numerical simulation and test based on ANSYS is adopted to simulate the external restraint of welding fixture by applying and releasing the constraint of displacement and concentrated force. The supporting force and clamping force of the workpiece are obtained by contact analysis. Finally, according to the designed test platform, three different multi-body coupling models are established to simulate the clamping action between the fixture and the workpiece, and the distribution of von Mises stress and XY direction stress in the upper and lower side plates of the workpiece are analyzed, as well as the forming mechanism of the X-Y direction deformation. Because the simulation results of external restraint model need to be tested and verified, a test platform for measuring the deformation of rotating arc angle welding workpiece has been set up in the laboratory. The angle deformation and bending deformation of the workpiece are obtained by measuring the corresponding measuring points of the lower side plate of the workpiece by using the special gauge for angle welding, and the numerical simulation platform for the thermal stress of the rotary arc angle welding is improved and improved.
【学位授予单位】：南昌大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG44

【参考文献】