基于接触特性的螺栓结合部动力学建模、参数识别和应用

发布时间：2018-12-06 15:57

【摘要】：数字仿真技术可以大幅度的缩短产品的设计周期,提高产品的竞争力,目前已经广泛应用于机床、飞机、汽车等工业设计相关领域中。建立准确的数学模型是实现精确仿真分析的基础。目前,对于单个零件的数字仿真分析的精确已经能够满足工程应用的要求。然而,对于系统级别的仿真分析的结果依旧有较大误差。其中一个重要的原因在于,现有的有限元软件不能很好的处理结构连接部分的性能。本文针对常用的螺栓连接结构的动力学建模、参数识别方法以及应用展开研究,主要研究内容包括以下几点:首先,建立了一种包含几何参数及物理参数的螺栓结合部虚拟材料动力学模型。通过对不同板厚、加工方式、螺栓直径及螺栓预紧力时的连接表面接触区域测试,分析了螺栓连接的影响因素对虚拟材料模型参数的影响程度。在大量实验的基础上,采用克里金法,建立了虚拟材料几何模型参数库,并基于matlab的GUI编辑了结合部几何参数库的管理界面。根据加工表面的纹理方向及接触情况,建立了虚拟材料的物理参数模型。其次,针对MAC值灵敏度计算耗时的问题,提出了一种基于拉格朗日函数的直接算法,将特征方程和正则化条件作为约束引入MAC值的计算公式中以构造拉格朗日函数,当求解出拉格朗日算子之后,可以直接计算MAC值对不同设计参数的一阶灵敏度及二阶灵敏度。并通过仿真算例分析,验证了直接法的准确性以及高效性。由于MAC值广泛应用于模型修正、参数识别等工程相关领域,直接法可以大幅度提高参数识别的效率。再次,根据本文提出的螺栓结合部虚拟材料动力学模型,提出了采用模态参数及频响函数相结合的方法进行虚拟材料物理参数的识别。在物理参数识别的过程中,根据修正方程的特性,提出了一种新的加权方法,显著减少修正方程的条件数,且方便的进行迭代运算。通过算例仿真,对参数识别方法的有效性进行了验证。设计螺栓连接的实验试件,进行螺栓结合部物理参数识别,并验证了参数识别结果的有效性。最后,围绕螺栓结合部模型的应用问题,构建了结合部模型物理参数库。根据不同的螺栓预紧力、粗糙度、加工方式以及接触方式,设计一系列的实验,并进行螺栓结合部虚拟材料模型物理参数的识别。在此基础上完成了机床螺栓结合部虚拟材料模型参数库的建立,并编写了相应的数据管理界面。最后,将螺栓结合部模型参数库应用于多螺栓长梁结构、框架结构及YK31320数控滚齿机的床身与底座连接结构,验证了模型库的有效性。全文主要围绕螺栓固定结合部展开:建立了螺栓结合部的动力学模型,提出了相应的参数识别方法,并构建了动力学模型参数库,实现了螺栓结合部的快速建模、分析以及应用。
[Abstract]:Digital simulation technology can greatly shorten the product design cycle, improve the competitiveness of the product, has been widely used in machine tools, aircraft, automobile and other industrial design related fields. Establishing accurate mathematical model is the basis of accurate simulation analysis. At present, the accuracy of digital simulation analysis for a single part can meet the requirements of engineering applications. However, there are still large errors in the results of system-level simulation analysis. One of the important reasons is that the existing finite element software can not deal well with the performance of the structural joint. In this paper, the dynamic modeling, parameter identification method and application of bolted connection structures are studied. The main research contents are as follows: first, A kind of virtual material dynamics model of bolt joint including geometric and physical parameters is established. The influence factors of bolt connection on the parameters of virtual material model were analyzed by testing the contact area of the connection surface with different plate thickness, processing method, bolt diameter and bolt preload force. On the basis of a large number of experiments, the virtual material geometric model parameter library is established by using Kriging method, and the management interface of the joint geometric parameter library is edited based on GUI of matlab. According to the texture direction and contact condition of the machined surface, the physical parameter model of the virtual material is established. Secondly, a direct algorithm based on Lagrange function is proposed to solve the problem of time-consuming calculation of MAC value sensitivity. The characteristic equation and regularization condition are introduced into the calculation formula of MAC value to construct the Lagrange function. When the Lagrange operator is solved, the first order and second order sensitivity of MAC value to different design parameters can be directly calculated. The accuracy and efficiency of the direct method are verified by simulation examples. Because the MAC value is widely used in engineering related fields such as model modification and parameter identification, the direct method can greatly improve the efficiency of parameter identification. Thirdly, according to the dynamic model of bolt joint virtual material, a method combining modal parameters and frequency response function is proposed to identify the physical parameters of virtual material. In the process of physical parameter identification, according to the characteristics of the modified equation, a new weighting method is proposed, which can significantly reduce the condition number of the modified equation, and is convenient for iterative operation. The effectiveness of the parameter identification method is verified by an example. The physical parameters of the bolt joint are identified and the validity of the parameter identification results are verified. Finally, the physical parameter library of bolt joint model is constructed around the application of bolt joint model. According to different bolt pre-tightening force, roughness, machining method and contact mode, a series of experiments are designed, and the physical parameters of the bolt joint virtual material model are identified. On this basis, the virtual material model parameter library of machine tool bolt joint is established, and the corresponding data management interface is compiled. Finally, the bolted joint model parameter library is applied to the multi-bolt long-beam structure, the frame structure and the connecting structure between the bed and the base of the YK31320 CNC hobbing machine, which verifies the validity of the model base. This paper mainly focuses on the bolt fixed joint: the dynamic model of the bolt joint is established, the corresponding parameter identification method is put forward, and the dynamic model parameter library is constructed, which realizes the fast modeling of the bolt joint. Analysis and application.
【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TH131.3

【相似文献】