节点文献
曲轴圆角强化滚压相关理论及关键技术研究
Study on Strengthen Theory and Key Technologies of Crankshaft Fillet Rolling
【作者】 张锐;
【导师】 胡荣强;
【作者基本信息】 武汉理工大学 , 交通信息工程与控制, 2010, 博士
【摘要】 曲轴是汽车发动机中最重要、承受交变载荷最大的零件。曲轴的主要破坏形式是弯曲疲劳破坏,其疲劳裂纹常常起源于曲柄至轴颈的过渡圆角处。圆角滚压可以大幅提高球墨铸铁曲轴的疲劳强度,鉴于球墨铸铁的密度小于锻钢,以球墨铸铁曲轴代替锻钢曲轴可以节省大量金属材料,降低成本。但是曲轴滚压的关键设备及技术一直由外国公司把持,国内销售价格居高不下,所以研究曲轴圆角滚压的关键技术并研制相关设备,具有重大的实际应用价值。本文介绍了曲轴滚压、摆差测试和校直工艺过程,分析了影响滚压强化效果的工艺参数。通过有限元数值模拟计算,详细分析了在不同控制加载方式、滚压圈数、滚压力、沉割槽等情况下,曲轴滚压强化效果,得出了应力、应变的分布规律,为滚压工艺优化提供了理论参照。在Windows环境下利用VC编程进行了压力闭环控制,建立了被控对象神经网络模型,构建了内模控制器,有效的抑制了动钳压力波动。论述了一种通过检测下压量来衡量残余应力的大小的检验方法,提出了将恒力滚压机通过简单的方式改造为变力滚压机的方法。建立了摆差测试数学模型,研制了基于LVDT的摆差测试硬件系统。当被测工件中心线和回转中心线不重合时,采用了矢量合成的数据处理方法以消除偏差。采用最小二乘法修正了摆差相位,采用了矢量模型变换方法以处理摆差数据,将顶尖测试基准下的结果变换为V型铁基准下的理论测试结果,在实践中获得了与曲轴实际变形手工测试基本一致的结果。利用最小二乘支持向量机结合多分类方法,构建了曲轴校直位置分类器。利用BP神经网络建立了校直专家决策数学模型,用数据融合的方法将多种决策下的控制量合成输出,在实用中取得了较好的效果。设计了脉冲方向型电机系统的接口电路,研制了具有自主知识产权的小型数控系统和PLC系统。制订了工控机IPC到数控系统的通信协议,在VC下编程实现了IPC与PLC系统间的通信,利用约定的存储单元实现了工作状态的同步控制。研究了曲轴滚压后的显微组织变化,找出了曲轴强化微观组织机理。滚压工艺使基体组织致密并且改善了应力集中现象,同时使滚压层晶内位错密度升高导致曲轴强度提高。测试了滚压部位的应力,结果表明在轴向和切向均产生了残余压应力,越靠近表层产生的残余压应力越大,这种残余压应力增强了曲轴疲劳强度。总之,文中采用有限元仿真、理论研究、样本学习和实验检验等等方法,研究了曲轴圆角强化、测试和校直工艺的关键技术。并成功研制出自动化程度较高的曲轴圆角滚压强化机床,填补了国内空白,占领了部分市场。
【Abstract】 Automotive engine crankshaft is the part of the most important and bear the largest load.The main destruction form of the crankshaft is bending and fatigue damage, and the origin of fatigue cracks on the stress concentration at the transition from the crank to the crankshaft journal fillet. Fillet rolling can significantly improve the fatigue strength of ductile iron crankshaft, because of the density of ductile iron is less than that of cast steel, a large number of metallic materials could been saved if ductile iron crankshaft is taken the place of cast steel crankshaft. Rolling process is been need to product the ductile iron crankshaft, while the crankshaft rolling equipment and market have been dominated by foreign companies,to study the key technologies on crankshaft fillet rolling and develope the related equipment have great significance in practical applications.The crankshaft rolling, runout testing and straightening processes were introduced in this paper. By FEM numerical simulation calculation, in different circumstances such as controlling load in different ways, rolling laps, rolling pressure, cutting groove, the crankshaft rolling and strengthening effect about stress and strain distribution were surveyed, a theoretical reference for the rolling process was provided.In Windows circumstance, VC programming for closed-loop control of pressure was programed. A neural network model of controlled object was established too. The internal model controller was constructed. The fluctuation of dynamic clamp pressure was suppressed effectively. A test method of measurement of the residual stress was discussed by detecting the depth of the depressed and a method that constant rolling machine would been transformed into inotropic rolling machine was proposed.A mathematical model of measurement deviation was established, simultaneously, the pendulum was developed based on LVDT differential test hardware system. When the axis of the measured workpiece does not coincide with that of rotary axis,the methods of the data vector synthesis should been used to eliminate bias.The differential phase of pendulum was amended by the method of least squares and the approach of vector model to transform data of runout was proposed. Finally, the top test results under the benchmark was transformed into the theory test results underⅤ-shaped iron, the actual deformation of the crankshaft basically is as same as the results of manual testing.By using the method of least squares support vector machine combining multiple classifiers, the crankshaft straightening position classifier was constructed. By the use of BP neural network, mathematical model of decision-making straightening expert was established, a variety of decision-making parameters were synthesized and output with the method of data treatment.A pulse and direction motors system interface circuits was designed, the small-scale numerical control systems and PLC systems were developed with independent intellectual property rights. IPC had developed into the numerical control system communication protocol, the communication between IPC and PLC was carried out with the VC program. By the use of the memory cell, a working state synchronization control was achieved.The microstructure of rolling crankshaft was investigated,the micro-mechanism of the crankshaft strengthen was identified. Rolling layer of grain was taken place local plastic deformation, so that the matrix densification and the stress concentration was improved. At the same time rolling layer of intragranular dislocation density was increased leading to the crankshaft strength increased. The stress test of rolling parts had been tested, the results shown that both the axial and tangential residual compressive stress was generated, the more close to the surface the larger value of residual compressive stress was generated.To Summary, By the using of the finite element simulation, theoretical studies, samples analyseing and experimental testing,the key technologies about crankshaft fillet enhanced tested and straightened process was studied. A high degree of automation Crankshaft Fillet Rolling Machine was successfully developed, occupying part of the domestic blank market.
【Key words】 fillet rolling; runout testing; FEM; straightening; BP neural network; support vector machine;