风扇单音噪声模态识别及其误差传递特性研究

Research of Mode Identification and Error Transfer Property on Fan Tonal Noise

关键词:  单音噪声  模态  阵列  叶轮机械  风扇

Key words:  Tonal noise  Mode  Array  Turbomachinery  Fan

摘要:

为了探究不同阵列对风扇单音噪声模态识别的影响，对径向耙和轴向传声器阵列在模态分解中的优缺点和误差传递特性进行了数值和实验研究。数值研究中通过在人造声场中加入随机扰动研究了不同阵列在径向模态分解中的误差传递特性，从而对两种阵列在径向模态分解方面的适用性和局限性进行评估。以此为指导，实验中通过旋转轴向阵列测量了风扇管道内顺流和逆流传播的模态波结构。结果表明中低频时径向耙式阵列误差较大，在模态截通频率处其传递函数的条件数大于100，而轴向阵列的条件数要小于10。轴向阵列在径向模态分解时，其精度强烈依赖于阵列轴向间距。实验结果表明管道内入射声波比反射声波能量高10dB左右。在入射声波中，转静干涉模态是其主导模态，而在反射声波中，模态能量分布规律不明显。

Abstract:

In order to analyze the influence on fan tonal noise mode decomposition with different arrays， the merits and drawbacks in mode decomposition using radial rakes and axial arrays and also the error transfer property were investigated in a numerical and experimental way. In the numerical study， the synthetic sound fields were superimposed by stochastic perturbation mode amplitude to analyze the error transfer characteristics of different arrays， the feasibility and limitation of arrays were evaluated subsequently. With above instruction， modal components propagating downstream and upstream were experimentally investigated using rotating axial arrays. Results show that error of radial rakes is larger in low and medium frequencies， the condition number of transfer function at mode cut-on frequencies is larger than 100， however， it is less than 10 in axial arrays. The precision of axial arrays is strongly dependent on the axial interval of the sensors in radial mode analysis. Experimental results show that the sound power of incident sound wave is 10dB higher than the reflected waves. Rotor-stator interaction modes dominate the incident mode waves， however， the general trend of the modal energy is inconclusive.