Comparison of Source Reconstruction Methods for Hybrid Aeroacoustic Predictions
Comparison of Source Reconstruction Methods for Hybrid Aeroacoustic Predictions
复制标题
混合气动声学预测的源重建方法比较
DOI：
--
复制
发表时间：
2013
期刊：
影响因子：
--
通讯作者：
W. Schröder
中科院分区：
文献类型：
--
作者：
G. Geiser;Daniel Marinc;W. Schröder
混合气动声学预测方法已经得到越来越多的普及，以克服流体力学和声学长度尺度之间的巨大差异时，亚音速流产生的噪声进行研究。这些预测的整体解决方案基本上取决于导致噪声发射的声源的准确表示。这些源通常不为人所知，因为它们与流体的湍流动力学相耦合。一种常见的方法是在感兴趣的小域上使用计算流体动力学来确定源，并且通过声学模型来计算较大域上的声场。这种耦合是敏感的，因为通常进行时间内插以减少适当的源表示所需的数据量。出现了插值对声学解的影响的问题，或者换句话说，什么样的插值公式是不影响声学结果所必需的。因此，在这项研究中，几个时间插值方法的质量进行评估，以确定适当的声源重建。首先，提出了一个通用的框架来设计插值滤波器，允许获得任意阶和可微的插值结果，并提供了一个可配置的通带优化。几个插值器进行了详细评估，特别照顾适当的声源重建。应用到亚音速湍流平面射流在雷诺数Re = 2000揭示了一个更好的源表示约束宽带优化插值。最小二乘优化的C 0-插值器与额外的泰勒约束，根据时间积分法的顺序显示出最好的结果。一个宽带误差减少原来是更重要的源重建相比，严格的收敛顺序。然而，为了实现整体解方案的期望收敛，时间积分器的泰勒阶应该至少是时间积分器的泰勒阶。埃尔米特插值被发现是太容易出错的不准确的数值近似所需的时间导数，即使有理论上的好处C1-可微的插值结果。
Hybrid aeroacoustic prediction methods have gained increasing popularity to overcome the large differences between the fluid mechanics and acoustics length scales when noise generated by subsonic flows is investigated. The overall solution of those predictions essentially depends on the accurate representation of the acoustic sources leading to noise emissions. The sources are not generally known, since they are coupled to the turbulent dynamics of the fluid. A common approach is to determine the sources using computational fluid dynamics on a small domain of interest and to compute by an acoustic model the sound field on a larger domain. This coupling is susceptible since usually an interpolation in time is conducted to reduce the amount of data required for a proper source representation. The question of the impact of the interpolation on the acoustic solution arises or in other words, what interpolation formulation is necessary to not impact the acoustic result. Therefore, in this study, the quality of several time-interpolation methods is assessed to determine an appropriate acoustic source reconstruction. First, a general framework to design interpolating filters is presented that allows to obtain interpolation results of arbitrary order and differentiability and offers a configurable passband optimization. Several interpolators are evaluated in detail with special care for proper acoustic source reconstruction. The application to a subsonic turbulent plane jet at Reynolds number Re = 2000 reveals a better source representation by constrained broadband optimized interpolation. Least-squares optimized C0-interpolators with additional Taylor constraints according to the order of the time integration method showed the best result. A broadband error reduction turned out to be more important for the source reconstruction when compared to a strict order of convergence. The Taylor order of the interpolator, however, should be at least the Taylor order of the temporal integrator to achieve the desired convergence of the overall solution scheme. Hermitian interpolation was found to be too error-prone to inaccuracies in the numerical approximation of the required temporal derivatives, even though having the theoretical benefit of C1-differentiability of the interpolation result.
飞机气动声控制的连续伴随优化程序研究
DOI：
10.1016/j.ijheatfluidflow.2012.07.005
发表时间：
2012
影响因子：
2.6
作者：
Daniel Marinc;Holger Foysi
通讯作者：
Holger Foysi