Direct numerical simulations of tonal noise generated by laminar flow past airfoils
Direct numerical simulations of tonal noise generated by laminar flow past airfoils
复制标题
DOI：
10.1016/j.jsv.2008.09.003
复制
发表时间：
2009-03
影响因子：
4.7
通讯作者：
R. Sandberg;L. E. Jones;N. Sandham;P. Joseph
中科院分区：
文献类型：
--
作者：
R. Sandberg;L. E. Jones;N. Sandham;P. Joseph
本文对M=0.4、雷诺数为Re= 50000时不同厚度、不同迎角的对称NACA翼型流动噪声进行了数值研究。采用直接数值模拟（DNS）直接计算近场流体力学和远场声场。然后，DNS数据被用于研究基于表面压差确定声调噪声辐射的方法，如Amiet的经典尾缘理论所做的那样，对受平均载荷影响的有限厚度翼型是否产生令人满意的结果。此外，还对Amiet表面压力跳变函数的精度进行了评价。总的来说，修正理论的阿米特似乎是适用于有限厚度翼型到中等发生率。然而，当将机翼厚度增加到12%弦时，即尾缘角为16.8°时，在受迫不稳定波的频率处，会发现入射压力和散射压力之间出现意想不到的相位变化。这种相位变化是由于流动以单独的尾迹频率在尾缘周围振荡。对于所研究的最大入射，Amiet的响应函数不能像在涡落频率下的零入射或小入射那样准确地预测总表面压差，从而导致对声压指向性和振幅的预测较差。此外，基于表面压差的翼型自噪声预测似乎并不普遍适用于较高的迎角，因为辐射声只是部分归因于经典的尾缘噪声机制。在这些情况下，似乎不能忽视流中的体积源。
A numerical investigation is presented of noise generated by flow past symmetric NACA airfoils with different thickness and at various angles of attack at M=0.4 and a Reynolds number based on chord of Re=50,000. Direct numerical simulations (DNS) are employed to directly compute both the near-field hydrodynamics and the far-field sound. The DNS data are then used to investigate whether the approach of determining tonal noise radiation based on the surface pressure difference, as done in the classical trailing-edge theory of Amiet, yields satisfactory results for finite thickness airfoils subject to mean loading effects. In addition, the accuracy of Amiet's surface pressure jump function is evaluated. Overall, the modified theory of Amiet appears to be suitable for finite thickness airfoils up to moderate incidence. However, when increasing the airfoil thickness to 12% chord, which corresponds to a trailing-edge angle of 16.8∘, an unexpected phase change between the incident and scattered pressure is found at the frequency of the forced instability waves. This phase change is attributed to the flow oscillating around the trailing edge at a separate wake frequency. For the largest incidence investigated, Amiet's response function does not predict the total surface pressure difference as accurately as for zero or small incidence at the vortex shedding frequency, resulting in a poor prediction of the directivity and amplitude of the acoustic pressure. Moreover, predicting the airfoil self-noise based on the surface pressure difference appears not to be generally applicable at higher angles of attack because the radiated sound is only partly due to classical trailing-edge noise mechanisms. In these cases, it appears as if volume sources in the flow cannot be neglected.