高速电主轴动态加载可靠性试验及其故障诊断研究

发布时间：2019-02-09 14:28

【摘要】：数控机床作为装备制造业的基础和关键设备,其性能与质量反映了一个国家制造业水平的高低。目前,可靠性水平是我国数控机床行业的技术瓶颈之一,因此,提高国产数控机床整机及其关键功能部件的可靠性水平势在必行。高速电主轴作为数控机床关键功能部件之一,通常是数控机床功率输出的核心部件,是目前制约我国数控机床整机可靠性水平的薄弱环节。因此,国产电主轴的可靠性水平亟待提高。本文以国产高速电主轴为研究对象,提出了考虑维修成本的电主轴FMECA方法,并根据现场可靠性试验数据,对电主轴进行故障分析;以机床实际工况的载荷数据为依据,编制了电主轴多维动态载荷谱,进而设计并开展了电主轴可靠性试验;为实现电主轴工作时的径向跳动及轴向窜动的状态监测,提出了一种基于加速度时-频域混合积分的电主轴振动监测方法;最后,提出了一种基于S变换改进阈值去噪的电主轴故障诊断方法,并进行了实例验证。本文的主要工作和研究成果如下:(1)论述了国内外数控机床电主轴可靠性技术的发展和研究现状,阐述了电主轴可靠性试验技术研究的背景及意义。对国内外电主轴的故障分析、电主轴静力加载及载荷谱加载的可靠性试验技术进行了探讨,综合评述了基于电主轴可靠性试验台的电主轴性能监测技术及故障诊断技术;在上述工作基础上,给出了本文关于电主轴可靠性试验及故障诊断研究的技术路线。(2)基于电主轴系统的组成结构与功能原理,对电主轴进行子系统划分。从电主轴故障模式分析入手,结合危害性分析的方法,对现场可靠性试验得到的故障数据进行了分析。针对传统危害度计算中存在的问题,即忽略了企业维修电主轴故障时付出的成本,提出一种考虑维修成本的危害度计算方法,并进行了危害度分析,得到了危害度最高的电主轴单元子系统,最终结果表明,该方法弥补了传统FMECA方法的不足,为电主轴可靠性试验、可靠性设计及可靠性增长提供了参考依据。(3)设计并开展了基于多维载荷谱动态加载的电主轴可靠性试验。鉴于以往载荷谱编制方法只能将切削力、扭矩、转速分别作谱,忽略了切削力、扭矩、转速间的对应关系,与实际工况不符的问题,提出电主轴的多维载荷谱编制思想。依据加工中心的现场载荷数据,分析并计算了切削力及切削扭矩。将切削力及切削扭矩按照轻载、轻中载、中载、中重载及重载,转速按低速、中速、高速分别进行分级,编制了包含转速、轴向力、径向力、扭矩及相对循环次数的多维载荷谱。最终,在电主轴试验台受载范围内设计并开展了电主轴的可靠性试验,包括空运转试验、静载试验及基于多维载荷谱动态加载的可靠性试验,并进行了电主轴可靠性评估。(4)针对数控机床切削作业时难以实时监测其径向跳动、轴向窜动的问题,提出了一种基于加速度传感器的时-频域混合积分的监测方法。对加速度信号进行高通滤波时域积分后,再经过低频截止频域积分,得到电主轴振动位移,即电主轴的径向跳动与轴向窜动。通过与Lab VIEW同步采集的数据进行积分误差检验表明,经时-频域混合积分法处理的结果与实测数据一致性较好。同时与简单二次时域积分法、简单二次频域积分法、低频截止频域积分法、多项式拟合时域积分法以及高通滤波时域积分法等五种方法进行了对比处理,试验结果表明,时-频域混合积分法得到的位移信号积分误差较小,更贴近实测位移信号,可以用于监测加工中心电主轴工作时的径向跳动及轴向窜动。(5)提出了一种基于S变换改进阈值去噪的电主轴故障诊断方法。在传统阈值算法的基础上,引入矫正参数,并结合S变换对故障特征频率进行提取,进而判断故障类型,最终实现电主轴故障诊断。基于高速电主轴可靠性试验的振动数据,采用S变换改进阈值去噪法对电主轴转子摩擦、不对中及机械松动等三种故障进行了诊断与识别。同时采用FFT、STFT以及HHT等三种故障诊断方法对故障数据进行处理,与S变换改进阈值去噪结果进行对比表明,FFT容易出现故障特征提取过度的问题,HHT则不能全面、精确地进行提取故障特征,而由于故障信号信噪比太差,STFT最不适用于电主轴故障诊断。对比分析结果证明了S变换改进阈值去噪方法在电主轴故障诊断中的优越性。
[Abstract]:As the base and key equipment of the equipment manufacturing industry, the performance and quality of the NC machine tool reflect the level of a country's manufacturing industry. At present, the reliability level is one of the technical bottleneck of the NC machine tool industry in China, so it is imperative to improve the reliability level of the whole machine and key functional parts of the home-made NC machine tool. As one of the key functional parts of NC machine, the high-speed electric spindle is usually the core part of the power output of the numerical control machine, which is the weak link that restricts the reliability level of the whole machine in China. Therefore, the reliability level of the domestic electric spindle needs to be improved. In this paper, using the domestic high-speed electric spindle as the research object, an electric spindle FMECA method considering the maintenance cost is put forward, and the fault analysis of the electric spindle is carried out according to the data of the on-site reliability test, the multi-dimensional dynamic load spectrum of the electric spindle is prepared based on the load data of the actual working condition of the machine tool, so as to design and carry out the reliability test of the electric spindle, and provides an electric spindle vibration monitoring method based on the acceleration time-frequency domain mixing integral to realize the state monitoring of the radial run-out and the axial movement during the work of the electric spindle, and finally, An electric spindle fault diagnosis method based on S-transform to improve the threshold de-noising is proposed, and the example verification is carried out. The main work and research results of this paper are as follows: (1) The development and research status of the reliability technology of the electric spindle of the NC machine at home and abroad are discussed, and the background and significance of the research on the reliability test technology of the electric spindle are described. The fault analysis of the main electric spindle at home and abroad, the static loading of the electric spindle and the reliability test technology of the load spectrum loading are discussed, and the performance monitoring technology and the fault diagnosis technology of the electric spindle based on the reliability test stand of the electric spindle are comprehensively reviewed. On the basis of the above work, The technical route of the reliability test and fault diagnosis of the electric spindle is given in this paper. (2) Based on the composition structure and functional principle of the electric spindle system, the sub-system of the electric spindle is divided. Based on the analysis of the failure mode of the electric spindle, the fault data obtained from the on-site reliability test is analyzed by means of the method of hazard analysis. Aiming at the problems existing in the calculation of the traditional hazard degree, that is, the cost of the failure of the maintenance electric spindle of the enterprise is ignored, a method for calculating the hazard degree of the maintenance cost is put forward, the harm degree analysis is carried out, the electric spindle unit sub-system with the highest harm degree is obtained, and the final result is that, The method overcomes the shortcomings of the traditional FMECA method, and provides a reference basis for the reliability test, the reliability design and the reliability growth of the electric spindle. (3) The reliability test of the electric spindle based on the dynamic loading of the multi-dimensional load spectrum is designed and carried out. In view of the previous load spectrum method, the cutting force, the torque and the rotating speed can be respectively used as the spectrum, the corresponding relation between the cutting force, the torque and the rotating speed is ignored, and the problem that the actual working condition is not in conformity with the actual working condition is ignored, and the multi-dimensional load spectrum preparation method of the electric spindle is put forward. According to the field load data of the machining center, the cutting force and cutting torque are analyzed and calculated. The cutting force and cutting torque are classified according to the light load, the light medium load, the medium load, the medium heavy load and the heavy load, and the rotating speed is respectively graded according to the low speed, the medium speed and the high speed, and the multi-dimensional load spectrum including the rotation speed, the axial force, the radial force, the torque and the relative cycle time is prepared. Finally, the reliability test of the electric spindle is carried out in the load range of the electric spindle test bed, including the air transfer test, the static load test and the reliability test based on the dynamic loading of the multi-dimensional load spectrum, and the reliability evaluation of the electric spindle is carried out. (4) It is difficult to monitor the radial runout and axial movement of the numerical control machine in real time, and a method for monitoring the time-frequency domain mixed integral based on the acceleration sensor is put forward. after the time-domain integral of the acceleration signal is high-pass filtered, the vibration displacement of the electric spindle is obtained through the low-frequency cut-off frequency domain integral, namely, the radial jump and the axial movement of the electric main shaft. The integration error test of the data collected with the Lab VIEW shows that the result of the time-domain mixed integration method is better with the measured data. At the same time, compared with the simple quadratic time-domain integral method, the simple quadratic frequency-domain integral method, the low-frequency cut-off frequency-domain integration method, the polynomial fitting time-domain integral method and the high-pass filtering time-domain integral method, the experimental results show that, and the displacement signal integration error obtained by the time-frequency domain mixing integration method is small, is closer to the measured displacement signal, and can be used for monitoring the radial runout and the axial movement of the machining center electric spindle during the work. (5) An electric spindle fault diagnosis method based on S-transform to improve the threshold de-noising is proposed. On the basis of the traditional threshold algorithm, the correction parameter is introduced, and the fault characteristic frequency is extracted in combination with the S transformation, and then the fault type is judged, and the fault diagnosis of the electric spindle is finally realized. Based on the vibration data of the reliability test of the high-speed electric spindle, the method of S-transform to improve the threshold de-noising method is used to diagnose and identify the three faults such as the friction, the non-centering and the mechanical loosening of the electric spindle. in that same time, three fault diagnosis methods such as FFT, STFT and HHT are adopted to process the fault data, and compared with the improvement threshold de-noising result of the S-transform, the problem that the FFT is prone to fault feature extraction is solved, and the HHT can not comprehensively and accurately extract the fault characteristic, Because the signal-to-noise ratio of the fault signal is too poor, the STFT is most suitable for the fault diagnosis of the electric spindle. The comparative analysis results show the superiority of the S-transform improved threshold de-noising method in the fault diagnosis of the electric spindle.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TG659

【参考文献】