Slat-Cove Noise Modeling: A Posteriori Analysis of Unsteady RANS Simulations

1. IntroductionA companion paper by Khorrami et al [_1demonstrates thefeasibility of simulating the (nominally) self-sustained,large-scale unsteadiness within the leading-edge slat-cove region of multi-element airfoils using unsteadyReynolds-Averaged Navier-Stokes (URANS)equations,provided that the turbulence production term in theunderlying two-equation turbulence model is switchedoff within the cove region. In conjunction with aFfowcsWilliams-Hawkings solver, the URANScomputations in ref. [1] were shown to capture thedominant portion of the acoustic spectrum attributed toslat noise, as well as reproducing the increased intensityof slat cove motions (and, correspondingly, far-fieldnoise as well) at the lower angles of attack. This paperexamines that simulation database, augmented byadditional simulations, with the objective of transitioningthis apparent success to aeroacoustic predictions in anengineering context. As a first step towards this goal, thesimulated flow and acoustic fields are compared withexperiments [2-51 and simplified analytical models [6'71.Rather intense near-field fluctuations in the simulatedflow are found to be associated with unsteady separationalong the slat bottom surface, relatively close to the slatcusp. Accuracy of the laminar-cove simulations in thisnear-wall region is raised to be an open issue. Theadjoint Green's function approach is also explored in anattempt to identify the most efficient noise sourcelocations.In studies on airframe noise, it is found that a primarysource of noise generation on high-lift devices involvesan interaction between large-scale energy-containingflow structures and the solid surfaces. This is especiallytrue near a flap side edge or within the slat cove of amulti-element airfoil configuration as indicated by thefavorable comparison between computational predictionsbased on the large-structure paradigm [1'8-91 and theexperimental measurements, either at the surface, I_°l or inthe far field E_°-121.Accordingly, the established approachin modern investigations I13j of airframe noise has been,first, to identify and model the unsteady disturbancessupported by relevant elements of the flow field and,then, to propagate this information to the far field viasome form of acoustic analogy. The part related to far-field propagation is conceptually straightforward,although its numerical implementation might requiresome thought and ingenuity as described in refs. [14]-[16]. Consequently, characterization of the unsteadyflow structures has become the crux of modeling thelarge class of airframe noise sources associated withseparated/free-shear flows.The required characterization of unsteady disturbancesmay be attempted at various levels of the modelinghierarchy, depending on the type of noise sourceinvolved and the specific purpose behind theinvestigation and the level of accuracy desired. Forinstance, the large-scale unsteadiness in a flap-side-edge+Aerospace Technologist, Computational Modeling and Simulation Branch, Senior Member AIAAAerospace Technologist, Computational Modeling and Simulation Branch, Associate Fellow AIAA**Visiting Research Professor, Fellow AIAACopyright © 2002 by the American Institute of Aeronautics and Astronautics, Inc. No copyright is asserted in the UnitedStates under Title 17, U.S. Code. The U.S. Government has a royalty-free license to exercise all rights under thecopyright claimed herein for Governmental Purposes. All other rights are reserved by the copyright owner.