节点文献

对旋轴流风机流固耦合特性研究及其基于非线性动力学的应用

Research on Fluid-Structure Interaction Characteristic of Counter Rotating Axial Flow Fan and Its Application Based on Nonlinear Dynamics

【作者】 周忠宁

【导师】 李意民;

【作者基本信息】 中国矿业大学 , 流体力学, 2009, 博士

【摘要】 叶轮机械的各种故障,如喘振、失速、叶片断裂等,与叶轮机械内各种流固耦合现象尤其是流体和叶轮叶片耦合有着直接的关系。因此对于叶轮机械内的非定常流场和叶片动力特性研究是流固耦合问题最重要的研究课题,对指导相关设备的设计和运行具有重要的意义。本文从实验研究、数值分析两个方面着手,以矿用对旋轴流风机为例全面分析了基于流固耦合考虑的三维流场特性与叶片动力特性,有针对性的给出了改善流场流动状况及叶片振动的方法。并在此基础上,根据对旋风机的流固耦合特性,进行了基于非线性理论的叶片振动非接触测量方法可行性研究、风机运行状态预测及故障判别等研究。本文首先通过流场试验与基于单传感器的叶片振动非接触测试试验获得了气体压力脉动信号、叶片振动信号,通过这些信号的频谱分析,揭示了对旋风机存在流场与叶片的耦合现象。其次进行了采用动力LES模型的风机全通道三维非定常流固耦合数值模拟。通过流场的实验与数值研究,得出了对旋风机耦合流场的一些非定常特征:前级转子的尾迹干扰及后级转子的势作用干扰;信号频率滑差;叶顶处压力脉动比叶跟、叶中的强烈;利用希尔伯特—黄变换对低频区域的压力脉动进行分析,发现了旋转失速、气流激振、喘振等非定常气动因素。数值模拟中揭示了叶片前缘二次流、叶片尾涡、通道涡、绕流、叶顶间隙流及间隙涡等流场固有非定常现象。针对试验与数值发现的流场叶顶区域流动复杂性,设计了叶片分离器改善叶顶部位的非定常流动,发现其对抑制对旋风机的波谷有一定作用。通过数值模拟得到了叶片表面压力分布及变形、应力等动力特性:叶片表面产生了较大的压力梯度,压力面的叶片前缘正压区域向叶片中部和后缘移动;叶片最大应力集中在根部且偏向叶片前缘,叶片根部应力受到流体质量和流动状态的双重影响;叶片的变形量变化与叶片应力变化趋势一致,随着流量的减少,变形位置由叶片顶端前缘向中部和后缘移动。为获得二级叶片更多的动力特性信息,本文首次同时将耦合流场与旋转预应力因素考虑到叶轮机械叶片的动力特性分析中。发现模态频率约10Hz的上升,模态振型发生了最大位移位置点转移、振动方向改变的状况;在叶片瞬态动力学分析中发现叶片的根部应力在较大的范围内波动。针对叶片动力特性的试验及数值模拟结果,提出了基于模态试验的叶片结构动力修改的方法,但这种方法的实际可操作性较差,同时指出在叶片设计过程中对风机叶片模态频率采用有限元分析,以避开叶片通过频率是改善叶片振动切实可行的方法。最后将级间流场的脉动时间序列信号首先进行了小波降噪处理,对叶片振动信号和降噪后的脉动信号进行相空间重构,获得了关联维、近似熵、L—Z复杂度的量化数据。发现叶片振动信号与气流脉动信号量化数据在不同工况下整体结构变化的趋势是一致的。根据对旋风机的流固耦合特性,提出了利用气动信号对风机叶片进行非接触测量与监测,并验证了这种方法的可行性;发现关联维数对失速信号是敏感的,可以用作判断风机旋转失速的特征量。同时根据量化数据与特征参数计算的结果与支持向量机结合建立了风机流场非线性系统的预测模型,进行了支持向量机的风机状态预测与诊断研究。

【Abstract】 This paper experimentally and numerically analyze the three dimensional flow field characteristics and the blade’s dynamic characteristics based on the consideration of fluid-solid coupling by using the counter rotating axial flow fan in mine as the example,then pointed a method to improve the flow condition and blade vibration. On the basis, according to the characteristic of the fluid-solid interaction in counter rotating axial flow fan, vibrated non-contact measuring technique was used to feasibility study, fan operation forecast and fault distinction based on the nonlinear theory.Firstly, taking the blade vibrated non-contact test based on the single sensor and the flow field test to obtain the gas pressure fluctuation and the blade vibration signal. Through spectrum analyzing, promulgated the phenomenon that the flow field coupled with blade in counter rotating axial flow fan.Then, the fan full channel three dimensional unsteady fluid-solid interaction numerical simulation was carried by using the dynamic LES model.Through the experiment and numerical research on the flow field, unsteady characteristics about the counter rotating axial flow fan’s coupling flow field and some dynamic characteristics including blade surface pressure distribution and distortion and stress were obtained in this paper.Finally the pulsating time sequence signal inter-stage flow field has first carried on wavelet de-noising processing, taking phase space reconstruction to the blade vibration signal and the de-noising pulsation signal, simultaneously obtained the quantitative data of correlation dimension, the approximate entropy, the L-Z complexity quantification data. Discovered the structure change tendency of the blade vibration signal and the air pulsation signal quantitative data is consistent under the different operation. Based on this, according to the fluid-solid interaction characteristic, proposed the method of blade non-contact measurement and monitor, and has confirmed this method feasibility; The correlation dimension to stall signal is sensitive which served as characteristic quantity to judge the rotating stall. Meanwhile has established the fan flow field nonlinear system forecast model according to combine the quantification data and the computation result of characteristic parameter and the support vector machines, has carried on the preliminary study of fan operation condition forecast and diagnosis.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络