混合陶瓷球轴承故障双冲击现象机理建模与性能退化指标评价

发布时间：2018-05-23 22:18

本文选题：混合陶瓷球轴承 + 轴承缺陷　；参考：《昆明理工大学》2016年硕士论文

【摘要】：滚动轴承作为旋转机械的关键基础部件,主要用于支撑转子旋转,且设备运行过程中大部分能量会传递到轴承上,进而导致轴承较易受损失效。随着近年来高速机械设备的发展,混合陶瓷球轴承因其自身的优势而在高端机床中得以广泛应用。然而,目前有关混合陶瓷球轴承的故障诊断技术及损伤程度评价方法却极为有限。通过分析故障轴承振动信号中双冲击特征实现剥落故障区长度的估算,为运行工况下轴承健康状态的判定及剩余寿命的预测提供一种可靠的方法。研究中以轴承外圈滚道单一剥落故障为例,在Hertz弹性理论的基础上深入分析滚动体通过不同尺寸大小缺陷时局部缺陷边缘与滚动体之间的弹性变形情况,且产生的冲击脉冲波形及幅值与缺陷形状和尺寸有着直接的关系,对于双冲击现象产生机理的研究也具有重要的学术研究价值。基于对滚动轴承内部缺陷故障诱发的非线性激励机理的研究,详细阐述了滚动体与外圈滚道间的接触位移及接触力的变化情况,并构建出相应的时变函数。通过在正常滚动轴承动力学模型的基础上提出时变位移与时变接触力相耦合的滚动轴承局部缺陷动力学模型,有效地克服单一函数时不变和时变位移激励缺陷模型难以描述局部剥落缺陷边缘接触变形过程及双冲击特征的体现,打破了目前基于单一函数缺陷模型的局限性。研究中利用龙格库塔数值积分法实现动力学微分方程的求解,经与轴承故障模拟实验测试分析结果对比,验证了故障轴承双冲击动力学模型的正确性。随着轴承的疲劳损伤积累,其性能从开始恶化至完全失效,需经历一个性能逐步退化的过程,因此需寻求能够实时监测轴承性能退化程度的特征量,便于及时更换轴承,防止轴承故障而导致设备异常失效的发生。本文分别从时域、频域及时频域三个特征域中分离提取能够表征轴承健康状态的特征指标,研究中重点分析了时频域中EMD、EEMD原理算法及性质,并引入EEMD能量熵作为时频域退化特征指标。由于所测轴承振动信号易受到其他零部件及外界噪声的干扰,故在特征提取前利用基于峭度最大准则的AR模型进行预白化处理,以消除信号中的确定成分。通过进行轴承加速疲劳寿命试验深入分析各特征指标对轴承性能退化过程的反映能力,尤其是初始性能退化的敏感性及全退化过程的一致性。
[Abstract]:As a key basic component of rotating machinery, rolling bearings are mainly used to support rotor rotation, and most of the energy will be transferred to the bearings during the operation of the equipment, which will lead to the loss of bearing efficiency. With the development of high-speed mechanical equipment, hybrid ceramic ball bearings are widely used in high-end machine tools because of their own advantages. However, the fault diagnosis techniques and damage evaluation methods for hybrid ceramic ball bearings are very limited. By analyzing the characteristics of double impact in the vibration signal of the fault bearing, the length of the spalling fault zone is estimated, which provides a reliable method for the determination of the bearing health state and the prediction of the remaining life under the operating condition. Taking the single spalling fault of the bearing outer ring raceway as an example, the elastic deformation between the local defect edge and the rolling body is deeply analyzed on the basis of Hertz elastic theory when the rolling body passes through different sizes and sizes of defects. The generated pulse waveform and amplitude are directly related to the shape and size of the defect, which is also of great academic value for the study of the mechanism of the double impact phenomenon. Based on the study of nonlinear excitation mechanism induced by internal defect fault of rolling bearing, the change of contact displacement and contact force between rolling body and outer ring raceway is described in detail, and the corresponding time-varying function is constructed. Based on the dynamic model of normal rolling bearing, a local defect dynamic model of rolling bearing is proposed, which is coupled with time-varying displacement and time-varying contact force. It can effectively overcome the limitation of single function time-invariant and time-varying displacement-induced defect model, which is difficult to describe the process of local flaking defect edge contact deformation and the expression of double impact characteristics, which breaks the limitation of the current defect model based on single function. In the study, the Runge-Kutta numerical integration method is used to solve the dynamic differential equation. The correctness of the double-impact dynamic model of the fault bearing is verified by comparing the results with the experimental results of bearing failure simulation. With the accumulation of fatigue damage, the performance of bearing deteriorates from beginning to complete failure, so it needs to go through a process of performance degradation step by step. Therefore, it is necessary to seek for the characteristic quantity which can monitor the performance degradation degree of bearing in real time so as to facilitate the timely replacement of bearing. Avoid bearing failure and cause abnormal failure of equipment. In this paper, the characteristic indexes which can represent the health state of bearings are separated and extracted from the time domain and the frequency domain respectively. The principle, algorithm and properties of EMD-EEMD in time-frequency domain are analyzed in detail. The energy entropy of EEMD is introduced as the characteristic index of degradation in time and frequency domain. Because the vibration signal of bearing is easily disturbed by other parts and external noise, the AR model based on kurtosis maximum criterion is used to pre-whiten the signal before feature extraction, so as to eliminate the definite components of the signal. Through the bearing accelerated fatigue life test, the ability of reflecting the degradation process of bearing performance, especially the sensitivity of initial performance degradation and the consistency of the whole degradation process, are analyzed deeply.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TH133.3

【相似文献】