Noise prediction for a turbulent jet using different hybrid methods
Noise prediction for a turbulent jet using different hybrid methods
复制标题
使用不同混合方法的湍流射流噪声预测
DOI：
10.1016/j.compfluid.2007.02.010
复制
发表时间：
2008
影响因子：
2.8
通讯作者：
P. Comte
中科院分区：
文献类型：
--
作者：
E. Gröschel;W. Schröder;P. Renze;M. Meinke;P. Comte
采用计算气动声学（CAA）方法，对马赫数为0.9、雷诺数为3600的单股冷喷流的声场进行了计算。射流计算的声场是由两个混合的方法，使用大涡模拟（LES）的流场和各种系统的方程的声场构建一个强大的，高效的，可靠的LES/CAA求解器。声学方程是频域中的Ffowcs Williams-Hawkings方程（FWH）和声学扰动方程（APE）。的FWH解决方案的方向性的数据窗口和径向和流向扩展的集成表面上的显着影响进行了讨论的长度。在相似的流动条件下，在近场的噪声特性的基础上，与现有的实验和数值结果的比较表明，APE系统的解决方案，以匹配的直接LES的结果比FWH方法更准确。APE的解决方案是不太容易受到的源项区域的大小比FWH方法的源表面的位置。与APE公式结合，LES域可以选择为小于FWH的ANEQUALITY，从而减少射流的计算成本。冷喷流噪声的APE系统中的主导源项被证明是兰姆矢量。
The acoustic field of a cold single stream jet at Mach number 0.9 and Reynolds number 3600 is determined via computational aeroacoustics (CAA) methods. The jet computation of the acoustical field is performed by two hybrid approaches using a large-eddy simulation (LES) for the flow field and various systems of equations for the acoustical field to construct a robust, efficient, and reliable LES/CAA solver. The acoustic equations are the Ffowcs Williams–Hawkings equation (FWH) in the frequency domain and the acoustic perturbation equations (APE). The pronounced impact of the data windowing and the radial and streamwise extension of the integration surface on the directivity of the FWH solution is discussed at length. The comparison with available experimental and numerical results at similar flow conditions based on the noise characteristics in the near field shows the solution of the APE system to match the results of the direct LES more accurately than the FWH approach. The APE solution is less susceptible to the size of the source term region than the FWH approach to the location of the source surface. In conjunction with the APE formulation the LES domain can be chosen smaller than for the FWH ansatz resulting in less computational cost for the jet flow. The dominant source term in the APE system for cold jet noise is shown to be the Lamb vector.