【作者】 王春瑞；

【导师】 岳林；

【作者基本信息】 南京航空航天大学 ， 机械设计及理论， 2011， 硕士

【摘要】 轴流压气机旋转失速是航空发动机常见气动不稳定现象之一。识别轴流压气机失速信号特征对于认识航空发动机失速的产生和发展机理以及抑制航空发动机失速问题并对其进行主动控制具有非常重要的意义。以往的信号处理方法对失速前后特征的识别做出了很大贡献,但是鉴于失速特征产生和发展机理极其复杂,所以对失速的研究一直在不断深入。本文主要应用三种方法来对失速特征进行分析。具体工作如下:1.初步学习并讨论了轴流压气机的压力测试方案。并且根据测得脉动压力的信号特点设计了低通FIR滤波器。2.应用互相关分析和同频检测方法分析旋转失速特征如失速团数目、失速团传播速度。结果表明该方法受噪声干扰无法精确确定时延,只适用于时延较长的低速和中速轴流压气机。高速轴流压气机会出现失速特征频率高于转子频率若干倍的高频失速,而用互相关分析方法无法准确求取失速团在两个传感器中传播的时延,为了解决该问题本文利用同频检测原理求取出失速信号的特征频率的时延,来对压气机高频失速的失速团数目进行识别。3.将脉动信号的相位和幅值特征通过极坐标进行可视化,对轴流压气机的失速特征进行分析。在极坐标中展示周向压力波动,可以较准确、直观的观察到失速团的数目;通过极坐标图形化脉动压力波动分析流场压力从正常工况到失速工况的规律可以对压气机的失速先兆进行预判。经过大量的试验数据分析,这种方法在分析失速团数目方面较相关分析等方法更加准确、直观;另外,本文还对压气机喘振的时域特征和频域特征进行了一定的分析。4.利用Matlab中GUI界面开发了一个简单实用的分析软件。该软件具有单/多通道信号时、频域显示、信号滤波、信号谱分析、失速团数目识别等功能。经过测试表明该软件可准确、方便的用于轴流压气机的失速特征分析。更多还原

【Abstract】 Rotating stall of axial compressor are common phenomenon of aeroengine instability, so it’s of great significance to identify its signal characteristics so as to better understand the mechanism it generated and developed, and then be prevented efficiently. The previous signal processing methods had made great contributions in recognizing the stall characteristic, but they are not that enough for stall in the occurrence and development mechanism. This paper mainly introduced three methods to analyze the characteristics of stall, the details are listed as follows:1. By learning the basic principles of the sensor and the test program of compressor’s dynamic pressure before. The paper designs a FIR filter according to the characteristics of fluctuating pressure signal we tested, which is proved to be effective by using the simulation data.2. Identify the number of stall cell and its propagation velocity by cross-correlation and the same frequency principle. The results show that the method can not be accurately determined delay time, which Applies only to a longer delay low and medium speed axial flow compressor. However, the stall characteristic frequency will be several times higher than the frequecny of rotor in high-speed compressor, so the cross-crrelation method can’t calculate the time delay acculately. In order to solve this problem, we use the same frequency detection principle to obtain time delay and identify the number of stall cell.3. By drawing the phase and amplitude characteristics of pressure fluctuation during stall process in the polar coordinate, we get the characteristic of rotating stall in compressor by analyzing the characteristics in polar coordinate. We can observe the variation of pressure in circumference of compressor and get the number of stall cells accutely and intuitively. In addition, according to law of normal to rotating stall condition, we can also do pre-judgment to stall inception. After analyzing a large number of test data, and this method has been proved to be more convenient, efficient and intuitive.4. We develop a simple but practical analysis software by using the GUI interface of Matlab. This software includes some basic functions such as signal and multi channel signal show in time domain and frequency domain, signal filtering, signal spectral analysis and the number of stall cells identification and so on. It has been tested that this software is effective in analyzing the dynamic pressure signal of compressor.更多还原