【作者】 秦鹏；

【导师】 蔡萍；

【作者基本信息】 上海交通大学 ， 测试计量技术及仪器， 2007， 博士

【摘要】 旋转机械的动不平衡是引起振动和噪声、降低设备寿命和可靠性、制约产品质量性能的主要原因之一,也是旋转类产品生产、制造以及应用过程中必须解决的共性问题。研究高精度动不平衡测量方法是一项具有巨大效益和重要意义的技术。振动系统和电气测量系统是动平衡设备的核心组成部分,针对提高外悬式转子的动不平衡测量设备的性能,本研究在新型振动系统设计、不平衡信号的提取和参数测量等环节提出新方法,以实现能够降低最小可达剩余不平衡量,提高平面分离性能、重复性和长期稳定性的动不平衡测量。具体研究工作分为以下几个方面:1.对外悬式转子动不平衡测量中的常规振动系统的结构特点、力的关联效应的产生机理及其抑制措施以及平面分离性能进行了深入分析,以提高不平衡测量精度和长期稳定性为目的,指出新型振动系统的设计途径。2.从理论力学中的瞬时运动中心的概念出发,提出了一种全新的全柔性串接式两自由度振动系统,该振动系统结构上可以兼具常规悬臂梁结构平衡效率高和简支梁结构平面分离好的优点为一体,而且拥有稳定的振动中心和较大的测量平面间距。将柔性铰链作为弹性元件引入到该新型振动系统中来,其在传递不平衡力和运动的方向上具有较低的刚度、阻尼和较高的灵敏度,降低了最小可达剩余不平衡量。传感元件安装在同一径向测量平面内,减弱了环境因素造成的传感器性能差异,从而抑制了关联效应对动不平衡测量精度和长期稳定性的影响。3.建立了该振动系统两自由度摆动的动力学模型,得到其运动微分方程,进行了不平衡响应分析,获得求解校正质量的平面分离方程。依据能量法原理,推导了该振动系统的等效刚度、固有频率和灵敏度的理论定量表达式,同时指出了其影响因素。利用ANSYS软件建立了振动系统的有限元理论模型,对其进行了静、模态分析,证明了理论建模的正确性。采用有限元理论深入分析了柔性铰链几何参数、振动系统结构参数和固有频率、灵敏度以及应力分布之间的影响关系,在此分析基础上,得到各个系统参数的优化取值范围。最后,初步结合机械结构优化设计方法,确定了振动系统的有关参数。4.设计实现了高性能的电气测量控制系统以确保实现高精度动不平衡测量。该电气测量控制系统以16位微处理器XC167为核心,主要包括两路压电传感器的电荷放大、选频滤波等信号调理电路;以提高相位分辨率和确定主轴转速、转向为目的的光电传感器选择及其后续电路设计;以及电机驱动控制电路。在此基础上,利用KeilC167实现了整套动不平衡测控程序。5.针对变频结构干扰和强噪声背景,分别采用改进经验模态分解和基于全通滤波器的自适应IIR格型陷波器的方法提取动不平衡信号。经验模态分解将振动信号自适应分解为有限多个由高频到低频排列的、正交的本征模态函数;利用自回归预测模型延拓信号端点,消除边界效应对提取的不平衡信号的影响,根据功率谱密度可以快速,有效的判断出代表基频信号的本征模态函数。对于基于全通滤波器的自适应IIR格型陷波器,提出了一种改进算法,该算法采用瞬时输入和号前一次瞬时输出信号的互相关来调节步长因子,提高了低信噪比条件下算法的收敛速度;引入归一化功率因子控制步长的变化量,使算法具有较好的鲁棒性能。实验结果证明了这两种方法的有效性。5.针对常规方法对不平衡信号参数测量需要整周期采样的限制,提出采用相位差校正技术进行动不平衡信号参数测量。该方法的采样频率不受待测信号频率影响,不需要整周期采样,算法实现方便,实时性好,选择旁瓣衰减较快的窗函数,能够消除能量泄露误差的影响,窗函数的长度选择为待测信号周期的整数倍,以消除谐波干扰的影响。实验表明结果该方法可以获得较高的测量精度。最后从平面分离性能、相位精度、重复性和最小可达剩余不平衡量方面进行了实验验证,结果表明该新型动不平衡测量系统的有效性。本研究取得的诸多成果可以为提高国有动不平衡测量设备性能提供理论指导和实践经验,同时也指出了未来需要进行的改进工作。更多还原

【Abstract】 Dynamic unbalancing of rotation machinery is one of sources of harmful vibration and noise, which affects life-span, reliability, and operation performance of the product. That is also a common problem in the course of manufacturing and application of rotation parts. It is meaningful to research high accuracy dynamic unbalancing measurement.Vibration system and electrical measurement system are the core components of the dynamic balancing equipment. To improve the performance of the dynamic unbalancing measurement equipment for outboard rotor, this paper presents some new methods during the course of unbalancing signal transmission, acquisition and processing, with the results that smaller minimum achievable residual unbalance, better plane separation performance and long-term stability is realized in unbalancing measurement. In brief, this research includes:1. Some points on outboard rotor unbalancing measurement, including the structure characteristics of the general vibration system, the mechanism of force relation effect, and the plane separation performance, are analyzed in depth. The design ideas for novel vibration system are proposed to improve measurement accuracy and long-tem stability.2.Starting from the concept of instantaneous motion center in theoretic mechanics, a novel fully compliant tandem vibration system with 2 DOF is presented. Its structure combines the advantage of the general overhanging beam structure with high balancing efficiency and the simple beam structure with good plane separation performance. Moreover, it has bigger distance between measuring planes and stable vibration center. Flexure hinges, which are considered as plastic element, are introduced to novel vibration system to provide higher sensitivity, low stiffness and damp in the direction of motion and force transmission in order that the minimum achievable residual unbalance is reduced. The transducers are installed on the same radial measuring plane, the performance changes in transducers caused by environment factor is reduced, the results for this vibration system in long-term operation is more accurate measurement with lower requirement for a second correction run.3. The kinematic model of vibration system is established and the kinematics differential equations are obtained. On the basis of analyzing unbalancing response, the plane separation equations are achieved. By energy method, the theoretic expressions of the equivalent stiffness, nature frequencies and sensitivity of vibration system are induced. Static and modal analyses of vibration system are performed by ANSYS finite element method to prove the reliability of the theoretical model. The influence of geometrical parameters of flexure hinges and structure parameters of vibration system on nature frequencies, sensitivity and stress distribution are researched in depth by FEM. Based on the above obtained results, an optimum design procedure is developed to determine vibration system.4. High performance electronic circuits is designed and implemented to achieve high accuracy dynamic unbalancing measurement. Includes: designing signal conditioning circuit (charge amplifier, selecting frequency and filter) for piezoelectric transducers, designing processing circuit for photodiode sensor to improve phase resolution and establish velocity and rotation direction of the principal axis, designing controlling circuit for motor based on 16-bit microprocessor XC167. The measuring and controlling programmer is realized by KeilC167.5. As the unbalancing signal being immerged under the strong background noises and varied frequency disturbances, the improved empirical modal decomposition method and an adaptive IIR lattice notch filter realized by all-pass filter are separately proposed to extract dynamic unbalancing signal. the original vibration signal can be adaptively decomposed into a definite number of orthogonal intrinsic mode functions arranged in order from high frequency to low frequency by empirical modal decomposition method. The auto-regression prediction model is introduced to eliminate the influence of end effects on balancing signal. The power spectral density is adopted to identify unbalancing signal from all intrinsic modal functions. On adaptive IIR lattice notch filter realized by all-pass filter, a novel algorithm is presented. The time-averaged estimation of cross correlation of the present instantaneous input signal and the past output signal is employed to update step-size. The convergence rate in a low SNR situation is considerably improved. The normalized power factor is introduced to control the variation of step-size in its steady-state bounds. This technique prevents algorithm from diverging due to the influence of biggish power input signal and improves the robustness of algorithm. The experiment results validate the effectiveness of the above two methods.6.The parameters of dynamic unbalancing signal, which are measured by the general methods, must be picked up by integral-period sampling. The phase difference correction method is presented to measure the parameters of random sampling unbalancing signal. The frequency of the measured signal has no influence on the sampling frequency when adopting this method and the measured signal needs not to be picked up by integral-period sampling. The proper windows function is selected to eliminate energy leakage error, and the length of the windows function should be the integer times of the period of unbalancing signal to eliminate harmonic disturbances. The experimental results validate the presented method can achieve high accuracy parameters measurement of dynamic balancing signal.Finally, the extensive experiments are made to prove the effectiveness of the presented dynamic unbalancing measurement system, including plane separation performance, phase accuracy, minimum achievable residual unbalance and repeatability. These achievements in paper can supply theoretical guidance and reference to improve the performance of national dynamic unbalancing measurement equipment.更多还原