Direct noise computation of adaptive control applied to a cavity flowSimulation directe de l'atténuation par contrôle adaptatif du rayonnement d'une cavité affleurée par un écoulement

The Large Eddy Simulation of closed-loop active flow control applied to a 3D cavity excited by a compressible airflow with a Mach number of 0.6 is presented. The control actuator is an idealized synthetic jet located at the upstream cavity edge, and the control function is supplied by a feedback LMS-type algorithm whose input is a pressure signal measured inside the cavity. The radiated sound, provided directly by the LES simulation, was shown to decrease substantially when active control was applied. A simultaneous reduction of the vertical velocity fluctuations in the shear layer was observed. The intensity of vortical structures inside the cavity was also reduced, although the general aspect of the recirculation zone was not modified. The direct noise computation technique, which supplies the pressure field by solving the fluid mechanics equations, is shown to constitute a powerful tool for studying active aeroacoustic noise control. To cite this article: O. Marsden et al., C. R. Mecanique 331 (2003).     

计算气动弹性若干研究进展

郭永怀和钱学森先生早在1946年提出了上临界马赫数的概念,即对于亚声速的二维无旋流 动,当来流速度达到下临界马赫数时开始出现声速. 稍增加来流速度,光滑无旋的亚、超声 速混合流动可以继续存在,理论上只有当来流速度达到上临界马赫数出现激波后, 光滑无旋流动才被破坏. 随后, 航空工程界先驱们为提高阻力发散马赫数,降低马赫数1附近的飞机 阻力, 为突破声障, 提出了超临界翼型设计技术,引进了后掠翼设计概念, 提出了跨声速面 积律理论,导致了20世纪军民用航空飞行器的大规模发展.随着计算机技术和计算方法的 进步,不同程度地简化流体控制方程的求解方法得到大发展.基于雷诺平均Navier-Stokes方程的计算流体力学已广泛应用于飞机性能评估、复杂流动机理分析.目前, 气动外形优化设计、气动/结构耦合干扰、气动噪声等多学科问题成为空气动力学的研究热点.该文介绍作者的团队近年来在计算气动弹性研究方面的若干进展,作为对郭永怀先生诞辰100周年的怀 念.      

激波-涡干扰声场的数值研究

采用高精度差分格式求解非定常可压缩Navier-Stokes方程,对激波－单涡／双涡相互干扰产生的声场进行了直接数值。详细研究了波－涡干扰声场结构的早期发展阶段,将激波－单涡的计算结果和相应实验进行 对比,并给出近场声压的衰减规律。在此基础上模拟较为复杂的激波－双涡干扰,给出不同旋涡旋转方向下的声场结构。     

Optimal low-dispersion low-dissipation LBM schemes for computational aeroacoustics

Lattice Boltmzann Methods (LBM) have been proved to be very effective methods for computational aeroacoustics (CAA), which have been used to capture the dynamics of weak acoustic fluctuations. In this paper, we propose a strategy to reduce the dispersive and disspative errors of the two-dimensional (2D) multi-relaxation-time lattice Boltzmann method (MRT-LBM). By presenting an effective algorithm, we obtain a uniform form of the linearized Navier–Stokes equations corresponding to the MRT-LBM in wave-number space. Using the matrix perturbation theory and the equivalent modified equation approach for finite difference methods, we propose a class of minimization problems to optimize the free-parameters in the MRT-LBM. We obtain this way a dispersion-relation-preserving LBM (DRP-LBM) to circumvent the minimized dispersion error of the MRT-LBM. The dissipation relation precision is also improved. And the stability of the MRT-LBM with the small bulk viscosity is guaranteed. Von Neuman analysis of the linearized MRT-LBM is performed to validate the optimized dispersion/dissipation relations considering monochromatic wave solutions. Meanwhile, dispersion and dissipation errors of the optimized MRT-LBM are quantitatively compared with the original MRT-LBM. Finally, some numerical simulations are carried out to assess the new optimized MRT-LBM schemes.     

Iterative optimization based on an objective functional in frequency-space with application to jet-noise cancellation

In many technical applications, like supersonic jets, noise with a characteristic spectrum including certain dominant frequencies (e.g. jet-screech) is prevalent, and the elimination of sharp peaks in the acoustic spectrum is the aim of active or passive flow/noise control efforts. A mathematical framework for the optimization of control strategies is introduced that uses a cost objective in frequency-space coupled to constraints in form of partial differential equations in the time domain. An iterative optimization scheme based on direct and adjoint equations arises, which has been validated on two examples, the one-dimensional Burgers equation and the two-dimensional compressible Navier–Stokes equations. In both cases, the iterative scheme has proven effective and efficient in targeting and removing specified frequency bands in the acoustic spectrum. It is expected that this technique will find use in acoustic and other applications where the elimination or suppression of distinct frequency components is desirable.     

Toward accurate hybrid prediction techniques for cavity flow noise applications

A large variety of hybrid computational aeroacoustics (CAA) approaches exist differing from each other in the way the source region is modeled, in the way the equations are used to compute the propagation of acoustic waves in a non-quiescent medium, and in the way the coupling between source and acoustic propagation regions is made. This paper makes a comparison between some commonly used numerical methods for aeroacoustic applications. The aerodynamically generated tonal noise by a flow over a 2D rectangular cavity is investigated. Two different cavities are studied. In the first cavity (L/D=4, M=0.5), the sound field is dominated by the cavity wake mode and its higher harmonics, originating from a periodical vortex shedding at the cavity leading edge. In the second cavity (L/D=2, M=0.6), shear-layer modes, due to flow-acoustic interaction phenomena, generate the major components in the noise spectrum. Source domain modeling is carried out using a second-order finite-volume large eddy simulation. Propagation equations, taking into account convection and refraction effects, are solved using high-order finite-difference schemes for the linearized Euler equations and the acoustic perturbation equations. Both schemes are compared with each other for various coupling methods between source region and acoustic region. Conventional acoustic analogies and Kirchhoff methods are rewritten for the various propagation equations and used to obtain near-field acoustic results. The accuracy of the various coupling methods in identifying the noise-generating mechanisms is evaluated. In this way, this paper provides more insight into the practical use of various hybrid CAA techniques to predict the aerodynamically generated sound field by a flow over rectangular cavities. Copyright © 2009 John Wiley & Sons, Ltd.     

空腔流激振荡发声的数值模拟研究

通过求解二维非定常雷诺平均Navier—Stokes方程，对亚音速情况下长宽比为L／D=2的空腔流动进行了计算气动声学数值模拟研究．湍流模型选取标准κ-ε模型，空间离散采用频散相关保持格式，时间积分采用低耗散低频散龙格库塔法，进出口远场边界采用以摄动解为基础的无反射边界条件，计算结果首先与Krishnamurty实验观测的密度场纹影图进行了对比，不论是在腔内流场分布还是腔外辐射指向性上，都与实验结果符合较好；然后与Rossiter公式进行了频率对比，其结果不但较为准确地捕捉到了振荡低阶模态的频率，而且也捕捉到了振荡高阶模态的频率特性，在此基础上研究了边界层厚度对振荡性质的影响，研究表明边界层厚度对振荡幅值影响较大，而且使振荡频率发生小幅偏移，最后探讨了振荡的发声机理，分析了振荡发声及声反馈的过程，发现了空腔前缘的二次发声现象。     

Computational aeroacoustics applications based on a discontinuous Galerkin method

caa simulation requires the calculation of the propagation of acoustic waves with low numerical dissipation and dispersion error, and to take into account complex geometries. To give, at the same time, an answer to both challenges, a Discontinuous Galerkin Method is developed for Computational AeroAcoustics. Euler's linearized equations are solved with the Discontinuous Galerkin Method using flux splitting technics. Boundary conditions are established for rigid wall, non-reflective boundary and imposed values. A first validation, for induct propagation is realized. Then, applications illustrate: the Chu and Kovasznay's decomposition of perturbation inside uniform flow in term of independent acoustic and rotational modes, Kelvin–Helmholtz instability and acoustic diffraction by an air wing. To cite this article: Ph. Delorme et al., C. R. Mecanique 333 (2005).     

离心叶轮气动声场的数值计算与分析

采用SIMPLEC算法对Ghost叶轮的三维非定常流场进行了数值模拟。利用计算所得流场结果并结合Lighthill和Lowson声学方程计算了由叶片表面非定常脉动力产生的气动噪声。计算结果表明:气动噪声的峰值主要集中在基频及其谐波附近;与静止的点声源相比,运动的点声源不仅使声场存在明显的多普勒效应,还会使声场的强度产生较大的变化;但对转速恒定的旋转点声源,加速度的变化对声场的影响可以忽略;从声场的分布来看,整个旋转叶轮可以看成是一个按简谐变化的偶极子源,数值计算结果与理论分析的结果吻合良好。     

Diffraction of acoustic waves on the leading edge of a flat plate in a supersonic flow

A scheme is proposed for calculating the intensity of the acoustic wave field generated by diffraction of a beam of acoustic waves on a sharp leading edge of a flat plate in a supersonic flow. This wave field is shown to be a functional of the mass-flow amplitude distribution in the acoustic field at the level of the plate surface upstream of the latter. This distribution can be found on the basis of measurements. The discontinuity of the normal-to-plate component of the velocity perturbation on the plate edge plays an important role in determining mass-flow fluctuations along the plate. At large distances from the leading edge of the plate, where the diffraction wave on the boundary-layer edge degenerates into longitudinal acoustic waves, the amplitude of mass-flow fluctuations decreases with increasing distance from the leading edge and depends on wave orientation.Translated from Prikladnaya Mekhanika i Tekhnicheskaya Fizika, Vol. 46, No. 2, pp. 64–70, March–April, 2005.