【作者】 唐兆；

【导师】 邬平波；

【作者基本信息】 西南交通大学 ， 载运工具运用工程， 2011， 博士

【摘要】 本文从机车车辆关键零部件疲劳仿真和可靠性分析的需求出发,总结了基于多体系统动力学仿真,有限元分析,疲劳数据处理和分析等工具进行机车车辆零部件疲劳仿真与分析的一般方法和处理流程。在自行研制的机车车辆应力试验数据处理系统和机车车辆疲劳仿真系统基础上,提出基于黏结集成算法的机车车辆疲劳仿真分析平台分布式架构体系,解决了试验与仿真数据的集成,不同计算任务之间的数据集成和数据交换,平台集成了现有相关的商业软件,具有友好开放的用户接口和良好的可扩展性,既提供方便的计算服务也保证了各软件的独立性,具有一定的推广应用价值。论文也对基于云计算理论的机车车辆疲劳强度仿真分析平台进行了初步讨论。针对机车车辆线路动应力试验和台架试验,综合常用的疲劳强度评定和疲劳寿命预测理论,利用C++和Delphi工具开发出了一套试验数据处理系统。系统具有与测试硬件无缝结合、试验报告自动生成、交互性好、可移植性强等特点；在传统三峰谷雨流计数算法的基础上,编程实现了一种针对大数据量的改进雨流计数方法,满足线路动应力试验对海量数据进行循环计数的需要。软件已在多条线路动应力试验数据处理中得到应用,论文以大秦线重载机车线路试验数据为基础,验证了软件的稳定性和分析结果的可靠性。机车车辆转向架的动应力时间历程获取是进行转向架疲劳寿命分析和预测最重要的环节。论文基于线路动应力试验,计算机仿真两种途径,编写了相应的工具软件。此外还提出了一种线路动应力试验和计算机模拟相结合的动应力获取方法,该方法以应变模态理论为基础,将试验数据引入到动应力数值模拟中来,为准确预测结构疲劳寿命奠定了基础。为将已贴片位置应力试验数据用于结构上不可贴片位置的应力大小预测,本文将试验和有限元分析相结合,基于逆方法,给出了用传感器响应方法(Sensor Response Method SRM)来预测不可贴片位置应力大小的具体方法,利用C++和Matlab,开发了SRM方法应力预测计算软件,将软件集成到仿真分析平台中,用于不可贴片位置应力的预测；并以CRH2转向架台架试验为例,验证了该方法的正确性和可靠性。更多还原

【Abstract】 For the actual need of fatigue strength and fatigue reliability analyses of the key parts in the vehicle industry, the paper summarized the vehicle fatigue simulation and analyses flow ultilizing MBS, FEM, Fatigue data reduction system and other software tools. On the basis of data analysis system of vehicle dynamic stress testing and vehicle fatigue simulation system which were developed by ourselves, the structure of vehicle fatigue simulation distribution platform based on gluing algorithm was set up. Data exchange and collaborative method among multiply disciplinary were discussed in detail. The platform integrated with commercial software has a friendly, open and easily removable interface and is valuable for practical use. The paper also discussed the application of cloud computing to the vehicle fatigue strength simulation platform.According to the line rolling stock dynamic stress test, based on the fatigue strength assessment and fatigue life prediction theory, we developed a set of experiment data processing system using c++and delphi. The system has the characteristics of seamless combination with test hardware; automatically generating test report; good interaction; great portability and so on. Based on the traditional tri-peak rain-flow counting algorithm, in view of a large amount of data, we realized the improvement of rain-flow counting method to satisfy the need of cycle count of the massive data in the line dynamic stress test. The software has been applied to the dynamic stress test data processes in some railway lines, such as the dynamic stress measurement data of heavy haul train for DaQing Line which was discussed in this paper. The stability of the software and the reliability of the analyzed results has been verified.Acquirement of the dynamic stress-time history is the key to fatigue life analysis and prediction. The software tools have been developed by means of testing in real field and numerical simulation. In addition, on the basis of stress/strain modals theory, a new method of obtaining dynamic stress-time history was put forward, which can ascertain the maximum stress/strain time history via other measured data and predict the fatigue life more precisely.On some specific positions, the dynamic stress-time history can’t be measured with strain gauge directly. Therefore, this paper presented a stress hybrid calculation method based on experimentation and numerical simulation. Based on inverse method named SRM(Sensor Response Method), the calculation formulas were deduced in detail in the paper. On the basis of these formulas, a software system using c++and Matlab was developed for vehicle bogie stress prediction. The strength experiment of CRH2bogie show that the stress can be obtained by the proposed method expediently and accurately.更多还原